IDS702 Final Project: HomeCredit Default Risk Dataset
William Huang
November 28, 2018
Project Objective
Home Credit Group, founded in Czech Republic a d headquartered in Netherlands, is a multinational non-bank financial company that provides consumer financial products, such as personal lending and credit card businesses. The company focuses primarily on people with no or little credit history. The project objective is to interpret how the information on the loan applications could affect the default risks of loan applicants.
Variable Descriptions
- TARGET: Target variable (1 - client with payment difficulties: he/she had late payment more than X days on at least one of the first Y installments of the loan in our sample, 0 - all other cases)
- CODE_GENDER: Gender of the client
- NAME_CONTRACT_TYPE: Identification if loan is cash or revolving
- AMT_INCOME_TOTAL: Income of the client
- FLAG_OWN_CAR: Flag if the client owns a car
- FLAG_OWN_REALTY: Flag if client owns a house or flat
- CNT_CHILDREN: Flag if client owns a house or flat
- AMT_CREDIT: Credit amount of the loan
- AMT_ANNUITY: Loan annuity
- AMT_GOODS_PRICE: For consumer loans it is the price of the goods for which the loan is given
- NAME_TYPE_SUITE: Who was accompanying client when he was applying for the loan
- NAME_INCOME_TYPE: Clients income type (businessman, working, maternity leave,…)
- NAME_EDUCATION_TYPE: Level of highest education the client achieved
- NAME_FAMILY_STATUS: Family status of the client
- NAME_HOUSING_TYPE: What is the housing situation of the client (renting, living with parents, …)
- REGION_POPULATION_RELATIVE: Normalized population of region where client lives (higher number means the client lives in more populated region)
- DAYS_BIRTH: Client’s age in days at the time of application
- DAYS_EMPLOYED: How many days before the application the person started current employment
- OWN_CAR_AGE: Age of client’s car
- OCCUPATION_TYPE: What kind of occupation does the client have
- CNT_FAM_MEMBERS: How many family members does client have
- ORGANIZATION_TYPE: Type of organization where client works
- AMT_REQ_CREDIT_BUREAU_DAY: Number of enquiries to Credit Bureau about the client one day before application (excluding one hour before application)
- AMT_REQ_CREDIT_BUREAU_WEEK: Number of enquiries to Credit Bureau about the client one week before application (excluding one day before application)
- AMT_REQ_CREDIT_BUREAU_MON: Number of enquiries to Credit Bureau about the client one month before application (excluding one week before application)
- AMT_REQ_CREDIT_BUREAU_QRT: Number of enquiries to Credit Bureau about the client 3 month before application (excluding one month before application)
- AMT_REQ_CREDIT_BUREAU_YEAR: Number of enquiries to Credit Bureau about the client one day year (excluding last 3 months before application)
Load Data and Data Pre-processing
As the original dataset contains over 200 independent variables, I picked the 23 variables for further analysis and interpretations based on my interest and work experience in this industry. Besides, I also added the following 6 variables together to improve the interpretability of my model: AMT_REQ_CREDIT_BUREAU_HOUR, AMT_REQ_CREDIT_BUREAU_DAY, AMT_REQ_CREDIT_BUREAU_WEEK, AMT_REQ_CREDIT_BUREAU_MON, AMT_REQ_CREDIT_BUREAU_QRT, AMT_REQ_CREDIT_BUREAU_YEAR. These variables represent the number of enquiries to Credit Bureau about the client one hour, day(excluding one hour before), week(excluding one day before), month(excluding one week before), quarter(excluding one month before) and year(excluding one quarter before) before application. As my goal is to construct a interpretation model, instead of a predictive model, it will enhance the interpretability of the model substantially by adding them together. On the other hand, if I intended to build a predictive model, I should keep them as 5 different variables.
#select interested variables
application_data_selected <- subset(application_data, select = c("TARGET","CODE_GENDER","NAME_CONTRACT_TYPE","AMT_INCOME_TOTAL","FLAG_OWN_CAR","FLAG_OWN_REALTY","CNT_CHILDREN","AMT_CREDIT", "AMT_ANNUITY", "AMT_GOODS_PRICE", "NAME_TYPE_SUITE","NAME_INCOME_TYPE","NAME_EDUCATION_TYPE","NAME_FAMILY_STATUS","NAME_HOUSING_TYPE","REGION_POPULATION_RELATIVE","DAYS_BIRTH","DAYS_EMPLOYED","OWN_CAR_AGE","OCCUPATION_TYPE","CNT_FAM_MEMBERS","ORGANIZATION_TYPE","AMT_REQ_CREDIT_BUREAU_HOUR","AMT_REQ_CREDIT_BUREAU_DAY","AMT_REQ_CREDIT_BUREAU_WEEK","AMT_REQ_CREDIT_BUREAU_MON","AMT_REQ_CREDIT_BUREAU_QRT","AMT_REQ_CREDIT_BUREAU_YEAR"))
#combine all the AMT_REQ_CREDIT_BUREAU variables together to compute the aggregated
#number of enquiries to Credit Bureau about the client one year before the application
application_data_selected$AMR_REQ_CREDIT_BUREAU_SUM <- application_data_selected$AMT_REQ_CREDIT_BUREAU_HOUR+application_data_selected$AMT_REQ_CREDIT_BUREAU_DAY+application_data_selected$AMT_REQ_CREDIT_BUREAU_MON+application_data_selected$AMT_REQ_CREDIT_BUREAU_QRT+application_data_selected$AMT_REQ_CREDIT_BUREAU_WEEK+application_data_selected$AMT_REQ_CREDIT_BUREAU_YEAR
#Drop AMT_REQ_CREDIT_BUREAU_HOUR, DAY, WEEK, MON, QRT, YEAR
application_data_selected <- application_data_selected %>%
dplyr::select(-(AMT_REQ_CREDIT_BUREAU_HOUR)) %>%
dplyr::select(-(AMT_REQ_CREDIT_BUREAU_DAY)) %>%
dplyr::select(-(AMT_REQ_CREDIT_BUREAU_WEEK)) %>%
dplyr::select(-(AMT_REQ_CREDIT_BUREAU_MON)) %>%
dplyr::select(-(AMT_REQ_CREDIT_BUREAU_QRT)) %>%
dplyr::select(-(AMT_REQ_CREDIT_BUREAU_YEAR))#check collinearity
num <- unlist(lapply(application_data_selected, is.numeric))
cor(cc(application_data_selected[,num]), use = "pair")## TARGET AMT_INCOME_TOTAL CNT_CHILDREN
## TARGET 1.000000000 -0.022374536 0.005679723
## AMT_INCOME_TOTAL -0.022374536 1.000000000 0.006490665
## CNT_CHILDREN 0.005679723 0.006490665 1.000000000
## AMT_CREDIT -0.035478230 0.324854819 -0.021083017
## AMT_ANNUITY -0.015778897 0.411376003 -0.001362760
## AMT_GOODS_PRICE -0.044389541 0.332695445 -0.024611028
## REGION_POPULATION_RELATIVE -0.036110371 0.167362370 -0.032117581
## DAYS_BIRTH 0.054896182 0.015586880 0.279672331
## DAYS_EMPLOYED -0.028323415 -0.103847372 -0.183114593
## OWN_CAR_AGE 0.037232346 -0.132171880 0.008950126
## CNT_FAM_MEMBERS -0.001250596 -0.001726251 0.914413945
## AMR_REQ_CREDIT_BUREAU_SUM 0.019143211 0.035277693 -0.029676739
## AMT_CREDIT AMT_ANNUITY AMT_GOODS_PRICE
## TARGET -0.03547823 -0.01577890 -0.04438954
## AMT_INCOME_TOTAL 0.32485482 0.41137600 0.33269544
## CNT_CHILDREN -0.02108302 -0.00136276 -0.02461103
## AMT_CREDIT 1.00000000 0.74634289 0.98718062
## AMT_ANNUITY 0.74634289 1.00000000 0.75161981
## AMT_GOODS_PRICE 0.98718062 0.75161981 1.00000000
## REGION_POPULATION_RELATIVE 0.09204365 0.11265566 0.09609678
## DAYS_BIRTH -0.11694218 -0.05319892 -0.11380798
## DAYS_EMPLOYED -0.01937100 -0.05245937 -0.01794351
## OWN_CAR_AGE -0.09280021 -0.09619776 -0.10185298
## CNT_FAM_MEMBERS 0.02020290 0.03046515 0.01779781
## AMR_REQ_CREDIT_BUREAU_SUM -0.01731215 0.01144510 -0.01839987
## REGION_POPULATION_RELATIVE DAYS_BIRTH
## TARGET -0.036110371 0.054896182
## AMT_INCOME_TOTAL 0.167362370 0.015586880
## CNT_CHILDREN -0.032117581 0.279672331
## AMT_CREDIT 0.092043647 -0.116942177
## AMT_ANNUITY 0.112655663 -0.053198919
## AMT_GOODS_PRICE 0.096096778 -0.113807976
## REGION_POPULATION_RELATIVE 1.000000000 -0.038847969
## DAYS_BIRTH -0.038847969 1.000000000
## DAYS_EMPLOYED 0.002191339 -0.508824142
## OWN_CAR_AGE -0.081436107 0.001110481
## CNT_FAM_MEMBERS -0.034209956 0.197954108
## AMR_REQ_CREDIT_BUREAU_SUM 0.025584835 -0.032038595
## DAYS_EMPLOYED OWN_CAR_AGE CNT_FAM_MEMBERS
## TARGET -0.028323415 0.037232346 -0.001250596
## AMT_INCOME_TOTAL -0.103847372 -0.132171880 -0.001726251
## CNT_CHILDREN -0.183114593 0.008950126 0.914413945
## AMT_CREDIT -0.019371002 -0.092800207 0.020202897
## AMT_ANNUITY -0.052459368 -0.096197763 0.030465151
## AMT_GOODS_PRICE -0.017943510 -0.101852976 0.017797807
## REGION_POPULATION_RELATIVE 0.002191339 -0.081436107 -0.034209956
## DAYS_BIRTH -0.508824142 0.001110481 0.197954108
## DAYS_EMPLOYED 1.000000000 0.031961957 -0.152888423
## OWN_CAR_AGE 0.031961957 1.000000000 -0.014499077
## CNT_FAM_MEMBERS -0.152888423 -0.014499077 1.000000000
## AMR_REQ_CREDIT_BUREAU_SUM 0.007042087 -0.026658773 -0.018385506
## AMR_REQ_CREDIT_BUREAU_SUM
## TARGET 0.019143211
## AMT_INCOME_TOTAL 0.035277693
## CNT_CHILDREN -0.029676739
## AMT_CREDIT -0.017312145
## AMT_ANNUITY 0.011445101
## AMT_GOODS_PRICE -0.018399870
## REGION_POPULATION_RELATIVE 0.025584835
## DAYS_BIRTH -0.032038595
## DAYS_EMPLOYED 0.007042087
## OWN_CAR_AGE -0.026658773
## CNT_FAM_MEMBERS -0.018385506
## AMR_REQ_CREDIT_BUREAU_SUM 1.000000000Based on the correlation table above, two pairs of independent variables have an extremely high correlation: (1) (AMT_CREDIT, AMT_GOODS_PRICE), corr = 0.987 AMT_GOODS_PRICE represents the goods price of good that client asked for on the previous application. AMT_CREDIT represents the final credit amount on the previous application. Through the definitions of these two variables, it’s clear that we could drop one of the variables.
- (
CNT_FAM_MEMBERS,CNT_CHILDREN), corr = 0.914CNT_FAM_mEMBERSrepresents how many family members clients haveCNT_CHILDRENrepresents how many children clients have Through the definitions of these two variables, it’s clear that we should drop one of the variables.
application_data_selected <- application_data_selected %>%
dplyr::select(-(AMT_GOODS_PRICE)) %>%
dplyr::select(-(CNT_CHILDREN))Here we should factorize CNT_FAM_MEMBERS, representing how many family members the loan applicant has, and AMT_REQ_CREDIT_BUREAU_SUM, representing number of enquiries to Credit Bureau about the client one year before submitting his loan application, because both variables are more like categorical rather than continuous, based on their definitions.
#factorize CNT_FAM_MEMBERS, AMR_REQ_CREDIT_BUREAU_SUM
application_data_selected$CNT_FAM_MEMBERS <- as.factor(application_data_selected$CNT_FAM_MEMBERS)
application_data_selected$AMR_REQ_CREDIT_BUREAU_SUM <- as.factor(application_data_selected$AMR_REQ_CREDIT_BUREAU_SUM)As the dataset is extremely large and my computer’s computing power doesn’t support this size of computations. Therefore, I decided to randomly select 10000 observations to run the following analysis. Besides, the random selection process should be able to represent the population.
set.seed(100)
application_data_sam <- application_data_selected[sample(nrow(application_data_selected), 10000), ]
summary(application_data_sam)## TARGET CODE_GENDER NAME_CONTRACT_TYPE AMT_INCOME_TOTAL
## Min. :0.0000 F :6556 Cash loans :9025 Min. : 25650
## 1st Qu.:0.0000 M :3444 Revolving loans: 975 1st Qu.: 112500
## Median :0.0000 XNA: 0 Median : 148500
## Mean :0.0829 Mean : 167765
## 3rd Qu.:0.0000 3rd Qu.: 202500
## Max. :1.0000 Max. :2250000
##
## FLAG_OWN_CAR FLAG_OWN_REALTY AMT_CREDIT AMT_ANNUITY
## N:6540 N:3070 Min. : 45000 Min. : 2844
## Y:3460 Y:6930 1st Qu.: 270000 1st Qu.: 16574
## Median : 521280 Median : 25047
## Mean : 599379 Mean : 27049
## 3rd Qu.: 808650 3rd Qu.: 34533
## Max. :2961000 Max. :133848
##
## NAME_TYPE_SUITE NAME_INCOME_TYPE
## Unaccompanied :8115 Working :5123
## Family :1278 Commercial associate:2335
## Spouse, partner: 374 Pensioner :1804
## Children : 118 State servant : 735
## Other_B : 50 Businessman : 1
## : 39 Student : 1
## (Other) : 26 (Other) : 1
## NAME_EDUCATION_TYPE NAME_FAMILY_STATUS
## Academic degree : 5 Civil marriage :1036
## Higher education :2414 Married :6320
## Incomplete higher : 305 Separated : 663
## Lower secondary : 114 Single / not married:1469
## Secondary / secondary special:7162 Unknown : 0
## Widow : 512
##
## NAME_HOUSING_TYPE REGION_POPULATION_RELATIVE DAYS_BIRTH
## Co-op apartment : 45 Min. :0.000938 Min. :-25186
## House / apartment :8863 1st Qu.:0.010006 1st Qu.:-19688
## Municipal apartment: 356 Median :0.018850 Median :-15820
## Office apartment : 80 Mean :0.020710 Mean :-16047
## Rented apartment : 179 3rd Qu.:0.028663 3rd Qu.:-12469
## With parents : 477 Max. :0.072508 Max. : -7721
##
## DAYS_EMPLOYED OWN_CAR_AGE OCCUPATION_TYPE CNT_FAM_MEMBERS
## Min. :-17546.0 Min. : 0.00 :3162 2 :5197
## 1st Qu.: -2729.0 1st Qu.: 5.00 Laborers :1793 1 :2186
## Median : -1181.5 Median : 9.00 Sales staff:1041 3 :1686
## Mean : 63952.6 Mean :11.97 Core staff : 878 4 : 814
## 3rd Qu.: -283.5 3rd Qu.:15.00 Managers : 667 5 : 101
## Max. :365243.0 Max. :65.00 Drivers : 623 6 : 11
## NA's :6540 (Other) :1836 (Other): 5
## ORGANIZATION_TYPE AMR_REQ_CREDIT_BUREAU_SUM
## Business Entity Type 3:2229 1 :1719
## XNA :1804 2 :1696
## Self-employed :1263 0 :1672
## Other : 529 3 :1304
## Business Entity Type 2: 354 4 : 906
## Government : 350 (Other):1376
## (Other) :3471 NA's :1327Data Description
Through the two tables below, we can see that the distributions of each variable in the original and new datasets are fairly similar. Besides, we also have some missing values and should consider missing value imputations.
summary(application_data_selected) #original datasets with selected variables## TARGET CODE_GENDER NAME_CONTRACT_TYPE
## Min. :0.00000 F :202448 Cash loans :278232
## 1st Qu.:0.00000 M :105059 Revolving loans: 29279
## Median :0.00000 XNA: 4
## Mean :0.08073
## 3rd Qu.:0.00000
## Max. :1.00000
##
## AMT_INCOME_TOTAL FLAG_OWN_CAR FLAG_OWN_REALTY AMT_CREDIT
## Min. : 25650 N:202924 N: 94199 Min. : 45000
## 1st Qu.: 112500 Y:104587 Y:213312 1st Qu.: 270000
## Median : 147150 Median : 513531
## Mean : 168798 Mean : 599026
## 3rd Qu.: 202500 3rd Qu.: 808650
## Max. :117000000 Max. :4050000
##
## AMT_ANNUITY NAME_TYPE_SUITE NAME_INCOME_TYPE
## Min. : 1616 Unaccompanied :248526 Working :158774
## 1st Qu.: 16524 Family : 40149 Commercial associate: 71617
## Median : 24903 Spouse, partner: 11370 Pensioner : 55362
## Mean : 27109 Children : 3267 State servant : 21703
## 3rd Qu.: 34596 Other_B : 1770 Unemployed : 22
## Max. :258026 : 1292 Student : 18
## NA's :12 (Other) : 1137 (Other) : 15
## NAME_EDUCATION_TYPE NAME_FAMILY_STATUS
## Academic degree : 164 Civil marriage : 29775
## Higher education : 74863 Married :196432
## Incomplete higher : 10277 Separated : 19770
## Lower secondary : 3816 Single / not married: 45444
## Secondary / secondary special:218391 Unknown : 2
## Widow : 16088
##
## NAME_HOUSING_TYPE REGION_POPULATION_RELATIVE DAYS_BIRTH
## Co-op apartment : 1122 Min. :0.00029 Min. :-25229
## House / apartment :272868 1st Qu.:0.01001 1st Qu.:-19682
## Municipal apartment: 11183 Median :0.01885 Median :-15750
## Office apartment : 2617 Mean :0.02087 Mean :-16037
## Rented apartment : 4881 3rd Qu.:0.02866 3rd Qu.:-12413
## With parents : 14840 Max. :0.07251 Max. : -7489
##
## DAYS_EMPLOYED OWN_CAR_AGE OCCUPATION_TYPE CNT_FAM_MEMBERS
## Min. :-17912 Min. : 0.00 :96391 2 :158357
## 1st Qu.: -2760 1st Qu.: 5.00 Laborers :55186 1 : 67847
## Median : -1213 Median : 9.00 Sales staff:32102 3 : 52601
## Mean : 63815 Mean :12.06 Core staff :27570 4 : 24697
## 3rd Qu.: -289 3rd Qu.:15.00 Managers :21371 5 : 3478
## Max. :365243 Max. :91.00 Drivers :18603 (Other): 529
## NA's :202929 (Other) :56288 NA's : 2
## ORGANIZATION_TYPE AMR_REQ_CREDIT_BUREAU_SUM
## Business Entity Type 3: 67992 1 :53914
## XNA : 55374 2 :51559
## Self-employed : 38412 0 :50911
## Other : 16683 3 :39380
## Medicine : 11193 4 :27241
## Business Entity Type 2: 10553 (Other):42987
## (Other) :107304 NA's :41519summary(application_data_sam) #the dataset after the random selection process## TARGET CODE_GENDER NAME_CONTRACT_TYPE AMT_INCOME_TOTAL
## Min. :0.0000 F :6556 Cash loans :9025 Min. : 25650
## 1st Qu.:0.0000 M :3444 Revolving loans: 975 1st Qu.: 112500
## Median :0.0000 XNA: 0 Median : 148500
## Mean :0.0829 Mean : 167765
## 3rd Qu.:0.0000 3rd Qu.: 202500
## Max. :1.0000 Max. :2250000
##
## FLAG_OWN_CAR FLAG_OWN_REALTY AMT_CREDIT AMT_ANNUITY
## N:6540 N:3070 Min. : 45000 Min. : 2844
## Y:3460 Y:6930 1st Qu.: 270000 1st Qu.: 16574
## Median : 521280 Median : 25047
## Mean : 599379 Mean : 27049
## 3rd Qu.: 808650 3rd Qu.: 34533
## Max. :2961000 Max. :133848
##
## NAME_TYPE_SUITE NAME_INCOME_TYPE
## Unaccompanied :8115 Working :5123
## Family :1278 Commercial associate:2335
## Spouse, partner: 374 Pensioner :1804
## Children : 118 State servant : 735
## Other_B : 50 Businessman : 1
## : 39 Student : 1
## (Other) : 26 (Other) : 1
## NAME_EDUCATION_TYPE NAME_FAMILY_STATUS
## Academic degree : 5 Civil marriage :1036
## Higher education :2414 Married :6320
## Incomplete higher : 305 Separated : 663
## Lower secondary : 114 Single / not married:1469
## Secondary / secondary special:7162 Unknown : 0
## Widow : 512
##
## NAME_HOUSING_TYPE REGION_POPULATION_RELATIVE DAYS_BIRTH
## Co-op apartment : 45 Min. :0.000938 Min. :-25186
## House / apartment :8863 1st Qu.:0.010006 1st Qu.:-19688
## Municipal apartment: 356 Median :0.018850 Median :-15820
## Office apartment : 80 Mean :0.020710 Mean :-16047
## Rented apartment : 179 3rd Qu.:0.028663 3rd Qu.:-12469
## With parents : 477 Max. :0.072508 Max. : -7721
##
## DAYS_EMPLOYED OWN_CAR_AGE OCCUPATION_TYPE CNT_FAM_MEMBERS
## Min. :-17546.0 Min. : 0.00 :3162 2 :5197
## 1st Qu.: -2729.0 1st Qu.: 5.00 Laborers :1793 1 :2186
## Median : -1181.5 Median : 9.00 Sales staff:1041 3 :1686
## Mean : 63952.6 Mean :11.97 Core staff : 878 4 : 814
## 3rd Qu.: -283.5 3rd Qu.:15.00 Managers : 667 5 : 101
## Max. :365243.0 Max. :65.00 Drivers : 623 6 : 11
## NA's :6540 (Other) :1836 (Other): 5
## ORGANIZATION_TYPE AMR_REQ_CREDIT_BUREAU_SUM
## Business Entity Type 3:2229 1 :1719
## XNA :1804 2 :1696
## Self-employed :1263 0 :1672
## Other : 529 3 :1304
## Business Entity Type 2: 354 4 : 906
## Government : 350 (Other):1376
## (Other) :3471 NA's :1327Missing values imputation
IWN_CAR_AGE and AMR_REQ_CREDOT_BUREAU_SUM are the two columns that contain missing values in the dataset. I used the mice package to conduct missing value imputation to generate complete datasets. For the imputation method, I chose cart, instead of the default method. I have tried to use the default method to impute missing values; however, it returned the following error “system is computationally singular”. The cause of the problem here could probably be the large number of unbalanced factor variables in the dataset. When these are turned intodummy variables there’s a high probability that one colum is a linear combination of another. As the default imputation methods involve linear regression, this results in a X matrix that cannot be inverted. Therefore, we consider to change the imputation method that is not stochastic, which require no X matrix inversion. (Reference: links)
#check pattern
md.pattern(application_data_sam)
## TARGET CODE_GENDER NAME_CONTRACT_TYPE AMT_INCOME_TOTAL FLAG_OWN_CAR
## 3069 1 1 1 1 1
## 5604 1 1 1 1 1
## 391 1 1 1 1 1
## 936 1 1 1 1 1
## 0 0 0 0 0
## FLAG_OWN_REALTY AMT_CREDIT AMT_ANNUITY NAME_TYPE_SUITE
## 3069 1 1 1 1
## 5604 1 1 1 1
## 391 1 1 1 1
## 936 1 1 1 1
## 0 0 0 0
## NAME_INCOME_TYPE NAME_EDUCATION_TYPE NAME_FAMILY_STATUS
## 3069 1 1 1
## 5604 1 1 1
## 391 1 1 1
## 936 1 1 1
## 0 0 0
## NAME_HOUSING_TYPE REGION_POPULATION_RELATIVE DAYS_BIRTH DAYS_EMPLOYED
## 3069 1 1 1 1
## 5604 1 1 1 1
## 391 1 1 1 1
## 936 1 1 1 1
## 0 0 0 0
## OCCUPATION_TYPE CNT_FAM_MEMBERS ORGANIZATION_TYPE
## 3069 1 1 1
## 5604 1 1 1
## 391 1 1 1
## 936 1 1 1
## 0 0 0
## AMR_REQ_CREDIT_BUREAU_SUM OWN_CAR_AGE
## 3069 1 1 0
## 5604 1 0 1
## 391 0 1 1
## 936 0 0 2
## 1327 6540 7867application_MI <- mice(application_data_sam, m = 10, method = "cart", seed = 8)## Warning: Number of logged events: 100Imputation Model Diagnostics
Based on the charts below, we are comfortable with our missing value imputations.
stripplot(application_MI, col=c("grey",mdc(2)),pch=c(1,20))
stripplot(application_MI, OWN_CAR_AGE~TARGET, col=c("grey",mdc(2)),pch=c(1,20), xlab = 'TARGET', ylab = "OWN_CAR_AGE")
stripplot(application_MI, AMR_REQ_CREDIT_BUREAU_SUM~TARGET, col=c("grey",mdc(2)),pch=c(1,20), xlab = 'TARGET', ylab = "AMT_REQ_CREDIT_BUREAU_DAY")
Posterior Predictive Check on two complete datasets
Both the histogram and boxplots look similiar for replica and complete datasets; therefore, we are confident about the quality of the imputation model.
application_ppcheck <- rbind(application_data_sam, application_data_sam)
application_ppcheck[10001:20000, apply(is.na(application_data_sam), any, MARGIN = 2)] <- NA
application_ppcheck_MI <- mice(application_ppcheck, m = 10, method = "cart", seed = 8)## Warning: Number of logged events: 100d1ppcheck <- mice::complete(application_ppcheck_MI, 1)
d2ppcheck <- mice::complete(application_ppcheck_MI, 2)#dataset1
par(mfrow = c(1,2))
boxplot(d1ppcheck$OWN_CAR_AGE[1:10000]~d1ppcheck$TARGET[1:10000], ylab="OWN_CAR_AGE", xlab="TARGET", main = "OWN_CAR_AGE vs TARGET completed data")
boxplot(d1ppcheck$OWN_CAR_AGE[10001:20000]~d1ppcheck$TARGET[10001:20000], ylab="OWN_CAR_AGE", xlab="TARGET", main = "OWN_CAR_AGE vs TARGET completed data")
#Should I treat it as continuous?
par(mfrow = c(2,1))
hist(as.numeric(d1ppcheck$AMR_REQ_CREDIT_BUREAU_SUM[1:10000]), xlab="AMR_REQ_CREDIT_BUREAU_SUM", main = "AMR_REQ_CREDIT_BUREAU_SUM complete data")
hist(as.numeric(d1ppcheck$AMR_REQ_CREDIT_BUREAU_SUM[10001:20000]), xlab="AMR_REQ_CREDIT_BUREAU_SUM", main = "AMR_REQ_CREDIT_BUREAU_SUM replicated data")
Regression Model
reg <- with(data = application_MI, glm(TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY
+ NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE
+ REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS
+ ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM),
family = binomial)
summary(pool(reg))## estimate
## (Intercept) 2.063259e-01
## CODE_GENDERM 4.352771e-02
## NAME_CONTRACT_TYPERevolving loans -5.055284e-02
## AMT_INCOME_TOTAL -5.315301e-08
## FLAG_OWN_CARY -3.081047e-02
## FLAG_OWN_REALTYY 5.206398e-03
## AMT_CREDIT -1.320492e-08
## AMT_ANNUITY -1.769943e-07
## NAME_TYPE_SUITEChildren -2.756176e-02
## NAME_TYPE_SUITEFamily -9.752861e-03
## NAME_TYPE_SUITEGroup of people -1.140243e-01
## NAME_TYPE_SUITEOther_A -6.362593e-02
## NAME_TYPE_SUITEOther_B -6.131117e-02
## NAME_TYPE_SUITESpouse, partner -9.428628e-03
## NAME_TYPE_SUITEUnaccompanied -1.514317e-02
## NAME_INCOME_TYPECommercial associate -1.027763e-01
## NAME_INCOME_TYPEPensioner -1.580174e-01
## NAME_INCOME_TYPEState servant -1.063998e-01
## NAME_INCOME_TYPEStudent -9.719079e-02
## NAME_INCOME_TYPEUnemployed -2.097608e-01
## NAME_INCOME_TYPEWorking -8.310605e-02
## NAME_EDUCATION_TYPEHigher education 3.705344e-02
## NAME_EDUCATION_TYPEIncomplete higher 5.441029e-02
## NAME_EDUCATION_TYPELower secondary 7.767959e-02
## NAME_EDUCATION_TYPESecondary / secondary special 5.559630e-02
## NAME_FAMILY_STATUSMarried -6.035944e-03
## NAME_FAMILY_STATUSSeparated -1.670707e-03
## NAME_FAMILY_STATUSSingle / not married 7.999912e-03
## NAME_FAMILY_STATUSWidow 1.727822e-02
## NAME_HOUSING_TYPEHouse / apartment 5.891942e-03
## NAME_HOUSING_TYPEMunicipal apartment -2.051035e-03
## NAME_HOUSING_TYPEOffice apartment -4.512994e-02
## NAME_HOUSING_TYPERented apartment 3.482144e-02
## NAME_HOUSING_TYPEWith parents 2.526804e-02
## REGION_POPULATION_RELATIVE -9.805655e-02
## DAYS_BIRTH 3.216099e-06
## DAYS_EMPLOYED 6.241027e-06
## OWN_CAR_AGE 1.037233e-04
## OCCUPATION_TYPEAccountants -1.463772e-02
## OCCUPATION_TYPECleaning staff -1.694824e-02
## OCCUPATION_TYPECooking staff 1.418259e-02
## OCCUPATION_TYPECore staff -5.402408e-03
## OCCUPATION_TYPEDrivers 1.028013e-02
## OCCUPATION_TYPEHigh skill tech staff -1.774165e-02
## OCCUPATION_TYPEHR staff -7.604784e-03
## OCCUPATION_TYPEIT staff -2.806577e-02
## OCCUPATION_TYPELaborers -4.816502e-03
## OCCUPATION_TYPELow-skill Laborers 2.681807e-02
## OCCUPATION_TYPEManagers -5.676823e-03
## OCCUPATION_TYPEMedicine staff -2.491503e-02
## OCCUPATION_TYPEPrivate service staff -6.379935e-03
## OCCUPATION_TYPERealty agents 7.190980e-02
## OCCUPATION_TYPESales staff -4.065929e-03
## OCCUPATION_TYPESecretaries 4.072782e-02
## OCCUPATION_TYPESecurity staff 4.084883e-02
## OCCUPATION_TYPEWaiters/barmen staff 1.002301e-01
## CNT_FAM_MEMBERS2 1.132116e-03
## CNT_FAM_MEMBERS3 -2.320473e-03
## CNT_FAM_MEMBERS4 -5.599006e-03
## CNT_FAM_MEMBERS5 3.401739e-02
## CNT_FAM_MEMBERS6 1.794479e-01
## CNT_FAM_MEMBERS7 8.906367e-02
## ORGANIZATION_TYPEAgriculture 4.382414e-03
## ORGANIZATION_TYPEBank 8.570364e-03
## ORGANIZATION_TYPEBusiness Entity Type 1 -1.214637e-03
## ORGANIZATION_TYPEBusiness Entity Type 2 3.669097e-02
## ORGANIZATION_TYPEBusiness Entity Type 3 8.953085e-03
## ORGANIZATION_TYPECleaning 3.671069e-02
## ORGANIZATION_TYPEConstruction 2.132647e-02
## ORGANIZATION_TYPECulture -5.861336e-02
## ORGANIZATION_TYPEElectricity -1.564123e-02
## ORGANIZATION_TYPEEmergency -2.077316e-02
## ORGANIZATION_TYPEGovernment 1.162231e-03
## ORGANIZATION_TYPEHotel -6.344843e-02
## ORGANIZATION_TYPEHousing 5.454882e-03
## ORGANIZATION_TYPEIndustry: type 1 -3.756790e-02
## ORGANIZATION_TYPEIndustry: type 11 1.558636e-02
## ORGANIZATION_TYPEIndustry: type 12 9.957355e-02
## ORGANIZATION_TYPEIndustry: type 13 -1.205705e-01
## ORGANIZATION_TYPEIndustry: type 2 -7.879773e-02
## ORGANIZATION_TYPEIndustry: type 3 4.168363e-02
## ORGANIZATION_TYPEIndustry: type 4 1.304441e-02
## ORGANIZATION_TYPEIndustry: type 5 1.303146e-02
## ORGANIZATION_TYPEIndustry: type 6 -7.803666e-02
## ORGANIZATION_TYPEIndustry: type 7 2.541053e-02
## ORGANIZATION_TYPEIndustry: type 9 3.122153e-02
## ORGANIZATION_TYPEInsurance -7.338810e-02
## ORGANIZATION_TYPEKindergarten -3.629599e-03
## ORGANIZATION_TYPELegal Services 1.982034e-02
## ORGANIZATION_TYPEMedicine 3.709787e-02
## ORGANIZATION_TYPEMilitary -3.655942e-02
## ORGANIZATION_TYPEMobile -9.156294e-02
## ORGANIZATION_TYPEOther 1.135235e-03
## ORGANIZATION_TYPEPolice -3.770840e-02
## ORGANIZATION_TYPEPostal -6.280312e-03
## ORGANIZATION_TYPERealtor 2.169619e-01
## ORGANIZATION_TYPEReligion -3.296117e-02
## ORGANIZATION_TYPERestaurant -1.767501e-02
## ORGANIZATION_TYPESchool 8.833369e-04
## ORGANIZATION_TYPESecurity -3.087008e-02
## ORGANIZATION_TYPESecurity Ministries -3.000705e-03
## ORGANIZATION_TYPESelf-employed 1.969723e-02
## ORGANIZATION_TYPEServices -1.091431e-02
## ORGANIZATION_TYPETelecom -2.065610e-02
## ORGANIZATION_TYPETrade: type 1 1.810593e-02
## ORGANIZATION_TYPETrade: type 2 -1.085085e-03
## ORGANIZATION_TYPETrade: type 3 -1.529725e-02
## ORGANIZATION_TYPETrade: type 4 -1.012080e-01
## ORGANIZATION_TYPETrade: type 6 -9.123168e-03
## ORGANIZATION_TYPETrade: type 7 3.726542e-02
## ORGANIZATION_TYPETransport: type 1 -7.505025e-02
## ORGANIZATION_TYPETransport: type 2 7.819173e-02
## ORGANIZATION_TYPETransport: type 3 1.384465e-01
## ORGANIZATION_TYPETransport: type 4 -3.459733e-02
## ORGANIZATION_TYPEUniversity -3.775120e-04
## ORGANIZATION_TYPEXNA -2.240731e+00
## AMR_REQ_CREDIT_BUREAU_SUM1 4.190379e-03
## AMR_REQ_CREDIT_BUREAU_SUM2 1.092473e-02
## AMR_REQ_CREDIT_BUREAU_SUM3 1.288864e-02
## AMR_REQ_CREDIT_BUREAU_SUM4 3.119740e-03
## AMR_REQ_CREDIT_BUREAU_SUM5 -7.012164e-03
## AMR_REQ_CREDIT_BUREAU_SUM6 2.503563e-02
## AMR_REQ_CREDIT_BUREAU_SUM7 1.588082e-02
## AMR_REQ_CREDIT_BUREAU_SUM8 5.843887e-02
## AMR_REQ_CREDIT_BUREAU_SUM9 1.691654e-02
## AMR_REQ_CREDIT_BUREAU_SUM10 -1.849285e-02
## AMR_REQ_CREDIT_BUREAU_SUM11 3.053713e-03
## AMR_REQ_CREDIT_BUREAU_SUM12 -4.140642e-02
## AMR_REQ_CREDIT_BUREAU_SUM13 -7.344946e-02
## AMR_REQ_CREDIT_BUREAU_SUM14 -1.746657e-02
## AMR_REQ_CREDIT_BUREAU_SUM15 -8.302970e-02
## AMR_REQ_CREDIT_BUREAU_SUM16 -2.768300e-02
## AMR_REQ_CREDIT_BUREAU_SUM17 -9.937823e-02
## AMR_REQ_CREDIT_BUREAU_SUM18 -7.341512e-02
## AMR_REQ_CREDIT_BUREAU_SUM19 -1.680945e-02
## AMR_REQ_CREDIT_BUREAU_SUM20 -5.056997e-02
## AMR_REQ_CREDIT_BUREAU_SUM28 -1.117664e-01
## std.error statistic
## (Intercept) 3.238824e-01 0.637039695
## CODE_GENDERM 7.349921e-03 5.922200671
## NAME_CONTRACT_TYPERevolving loans 9.804538e-03 -5.156065232
## AMT_INCOME_TOTAL 3.806233e-08 -1.396472747
## FLAG_OWN_CARY 6.335356e-03 -4.863258282
## FLAG_OWN_REALTYY 6.208820e-03 0.838548695
## AMT_CREDIT 1.107826e-08 -1.191966886
## AMT_ANNUITY 3.253000e-07 -0.544095497
## NAME_TYPE_SUITEChildren 5.078004e-02 -0.542767609
## NAME_TYPE_SUITEFamily 4.468996e-02 -0.218233848
## NAME_TYPE_SUITEGroup of people 1.298177e-01 -0.878341280
## NAME_TYPE_SUITEOther_A 7.422377e-02 -0.857217725
## NAME_TYPE_SUITEOther_B 5.868265e-02 -1.044792141
## NAME_TYPE_SUITESpouse, partner 4.628181e-02 -0.203722120
## NAME_TYPE_SUITEUnaccompanied 4.412372e-02 -0.343197977
## NAME_INCOME_TYPECommercial associate 2.844942e-01 -0.361259798
## NAME_INCOME_TYPEPensioner 3.946596e-01 -0.400389083
## NAME_INCOME_TYPEState servant 2.848237e-01 -0.373563697
## NAME_INCOME_TYPEStudent 3.901416e-01 -0.249116717
## NAME_INCOME_TYPEUnemployed 4.799008e-01 -0.437092064
## NAME_INCOME_TYPEWorking 2.846590e-01 -0.291949486
## NAME_EDUCATION_TYPEHigher education 1.222726e-01 0.303039512
## NAME_EDUCATION_TYPEIncomplete higher 1.231872e-01 0.441687961
## NAME_EDUCATION_TYPELower secondary 1.248090e-01 0.622387764
## NAME_EDUCATION_TYPESecondary / secondary special 1.221992e-01 0.454964697
## NAME_FAMILY_STATUSMarried 9.324405e-03 -0.647327578
## NAME_FAMILY_STATUSSeparated 1.701766e-02 -0.098174855
## NAME_FAMILY_STATUSSingle / not married 1.680871e-02 0.475938598
## NAME_FAMILY_STATUSWidow 2.034232e-02 0.849372761
## NAME_HOUSING_TYPEHouse / apartment 4.101925e-02 0.143638448
## NAME_HOUSING_TYPEMunicipal apartment 4.344237e-02 -0.047212774
## NAME_HOUSING_TYPEOffice apartment 5.133026e-02 -0.879207432
## NAME_HOUSING_TYPERented apartment 4.573693e-02 0.761341852
## NAME_HOUSING_TYPEWith parents 4.277699e-02 0.590692396
## REGION_POPULATION_RELATIVE 2.097866e-01 -0.467410822
## DAYS_BIRTH 9.715712e-07 3.310204273
## DAYS_EMPLOYED 1.476965e-06 4.225575963
## OWN_CAR_AGE 2.801423e-04 0.370252048
## OCCUPATION_TYPEAccountants 1.778898e-02 -0.822853488
## OCCUPATION_TYPECleaning staff 2.441296e-02 -0.694231084
## OCCUPATION_TYPECooking staff 2.101915e-02 0.674746192
## OCCUPATION_TYPECore staff 1.324128e-02 -0.407997419
## OCCUPATION_TYPEDrivers 1.444968e-02 0.711443845
## OCCUPATION_TYPEHigh skill tech staff 1.559747e-02 -1.137469734
## OCCUPATION_TYPEHR staff 7.360465e-02 -0.103319349
## OCCUPATION_TYPEIT staff 6.037217e-02 -0.464879335
## OCCUPATION_TYPELaborers 1.040000e-02 -0.463125263
## OCCUPATION_TYPELow-skill Laborers 3.512250e-02 0.763558124
## OCCUPATION_TYPEManagers 1.342852e-02 -0.422743750
## OCCUPATION_TYPEMedicine staff 2.288624e-02 -1.088646858
## OCCUPATION_TYPEPrivate service staff 3.046778e-02 -0.209399389
## OCCUPATION_TYPERealty agents 5.679346e-02 1.266163399
## OCCUPATION_TYPESales staff 1.227742e-02 -0.331171413
## OCCUPATION_TYPESecretaries 4.763194e-02 0.855052630
## OCCUPATION_TYPESecurity staff 2.237480e-02 1.825662323
## OCCUPATION_TYPEWaiters/barmen staff 4.716424e-02 2.125129651
## CNT_FAM_MEMBERS2 1.464121e-02 0.077323907
## CNT_FAM_MEMBERS3 1.639724e-02 -0.141516061
## CNT_FAM_MEMBERS4 1.817850e-02 -0.308001501
## CNT_FAM_MEMBERS5 3.137761e-02 1.084129310
## CNT_FAM_MEMBERS6 8.416156e-02 2.132183844
## CNT_FAM_MEMBERS7 1.237411e-01 0.719758220
## ORGANIZATION_TYPEAgriculture 7.110094e-02 0.061636519
## ORGANIZATION_TYPEBank 7.345444e-02 0.116675911
## ORGANIZATION_TYPEBusiness Entity Type 1 6.733533e-02 -0.018038620
## ORGANIZATION_TYPEBusiness Entity Type 2 6.608540e-02 0.555205375
## ORGANIZATION_TYPEBusiness Entity Type 3 6.467652e-02 0.138428667
## ORGANIZATION_TYPECleaning 1.114015e-01 0.329534866
## ORGANIZATION_TYPEConstruction 6.707111e-02 0.317968028
## ORGANIZATION_TYPECulture 9.999669e-02 -0.586153024
## ORGANIZATION_TYPEElectricity 8.055834e-02 -0.194160226
## ORGANIZATION_TYPEEmergency 8.792009e-02 -0.236273141
## ORGANIZATION_TYPEGovernment 6.604658e-02 0.017597149
## ORGANIZATION_TYPEHotel 8.257608e-02 -0.768363320
## ORGANIZATION_TYPEHousing 6.982835e-02 0.078118440
## ORGANIZATION_TYPEIndustry: type 1 7.841714e-02 -0.479077683
## ORGANIZATION_TYPEIndustry: type 11 7.098229e-02 0.219580928
## ORGANIZATION_TYPEIndustry: type 12 1.020855e-01 0.975393521
## ORGANIZATION_TYPEIndustry: type 13 2.036842e-01 -0.591948495
## ORGANIZATION_TYPEIndustry: type 2 9.731529e-02 -0.809715791
## ORGANIZATION_TYPEIndustry: type 3 6.914345e-02 0.602857174
## ORGANIZATION_TYPEIndustry: type 4 7.817763e-02 0.166856096
## ORGANIZATION_TYPEIndustry: type 5 8.878250e-02 0.146779649
## ORGANIZATION_TYPEIndustry: type 6 1.380145e-01 -0.565423688
## ORGANIZATION_TYPEIndustry: type 7 7.682659e-02 0.330751780
## ORGANIZATION_TYPEIndustry: type 9 6.945843e-02 0.449499499
## ORGANIZATION_TYPEInsurance 9.119419e-02 -0.804745374
## ORGANIZATION_TYPEKindergarten 6.709661e-02 -0.054095110
## ORGANIZATION_TYPELegal Services 9.761973e-02 0.203036166
## ORGANIZATION_TYPEMedicine 6.714712e-02 0.552486377
## ORGANIZATION_TYPEMilitary 7.111523e-02 -0.514086997
## ORGANIZATION_TYPEMobile 1.046480e-01 -0.874961544
## ORGANIZATION_TYPEOther 6.550997e-02 0.017329199
## ORGANIZATION_TYPEPolice 7.237334e-02 -0.521026129
## ORGANIZATION_TYPEPostal 7.224699e-02 -0.086928360
## ORGANIZATION_TYPERealtor 1.045732e-01 2.074736910
## ORGANIZATION_TYPEReligion 2.033270e-01 -0.162109178
## ORGANIZATION_TYPERestaurant 7.372632e-02 -0.239738171
## ORGANIZATION_TYPESchool 6.659657e-02 0.013263999
## ORGANIZATION_TYPESecurity 7.165784e-02 -0.430798349
## ORGANIZATION_TYPESecurity Ministries 7.303849e-02 -0.041083890
## ORGANIZATION_TYPESelf-employed 6.488590e-02 0.303567190
## ORGANIZATION_TYPEServices 7.689387e-02 -0.141939911
## ORGANIZATION_TYPETelecom 8.996517e-02 -0.229601124
## ORGANIZATION_TYPETrade: type 1 1.045225e-01 0.173225100
## ORGANIZATION_TYPETrade: type 2 7.420166e-02 -0.014623457
## ORGANIZATION_TYPETrade: type 3 6.905253e-02 -0.221530686
## ORGANIZATION_TYPETrade: type 4 2.036165e-01 -0.497052183
## ORGANIZATION_TYPETrade: type 6 8.610919e-02 -0.105948836
## ORGANIZATION_TYPETrade: type 7 6.674541e-02 0.558321819
## ORGANIZATION_TYPETransport: type 1 1.510872e-01 -0.496734785
## ORGANIZATION_TYPETransport: type 2 7.187849e-02 1.087832087
## ORGANIZATION_TYPETransport: type 3 7.668451e-02 1.805403139
## ORGANIZATION_TYPETransport: type 4 6.837164e-02 -0.506018664
## ORGANIZATION_TYPEUniversity 7.928623e-02 -0.004761382
## ORGANIZATION_TYPEXNA 6.089517e-01 -3.679653516
## AMR_REQ_CREDIT_BUREAU_SUM1 9.127409e-03 0.459098434
## AMR_REQ_CREDIT_BUREAU_SUM2 9.816924e-03 1.112846530
## AMR_REQ_CREDIT_BUREAU_SUM3 1.053364e-02 1.223569960
## AMR_REQ_CREDIT_BUREAU_SUM4 1.161135e-02 0.268680246
## AMR_REQ_CREDIT_BUREAU_SUM5 1.360600e-02 -0.515373121
## AMR_REQ_CREDIT_BUREAU_SUM6 1.725497e-02 1.450922280
## AMR_REQ_CREDIT_BUREAU_SUM7 1.898283e-02 0.836588629
## AMR_REQ_CREDIT_BUREAU_SUM8 2.739782e-02 2.132974918
## AMR_REQ_CREDIT_BUREAU_SUM9 3.608539e-02 0.468792048
## AMR_REQ_CREDIT_BUREAU_SUM10 4.468145e-02 -0.413882025
## AMR_REQ_CREDIT_BUREAU_SUM11 7.004005e-02 0.043599527
## AMR_REQ_CREDIT_BUREAU_SUM12 1.057255e-01 -0.391640785
## AMR_REQ_CREDIT_BUREAU_SUM13 1.157125e-01 -0.634758237
## AMR_REQ_CREDIT_BUREAU_SUM14 1.799169e-01 -0.097081323
## AMR_REQ_CREDIT_BUREAU_SUM15 1.321943e-01 -0.628088210
## AMR_REQ_CREDIT_BUREAU_SUM16 2.742893e-01 -0.100926295
## AMR_REQ_CREDIT_BUREAU_SUM17 2.726974e-01 -0.364426748
## AMR_REQ_CREDIT_BUREAU_SUM18 1.840078e-01 -0.398978291
## AMR_REQ_CREDIT_BUREAU_SUM19 2.746293e-01 -0.061207775
## AMR_REQ_CREDIT_BUREAU_SUM20 2.562267e-01 -0.197364150
## AMR_REQ_CREDIT_BUREAU_SUM28 2.599306e-01 -0.429985680
## df p.value
## (Intercept) 9842.13067 5.241138e-01
## CODE_GENDERM 9847.07017 3.282767e-09
## NAME_CONTRACT_TYPERevolving loans 9667.51517 2.570858e-07
## AMT_INCOME_TOTAL 9030.00715 1.626036e-01
## FLAG_OWN_CARY 9859.54073 1.172539e-06
## FLAG_OWN_REALTYY 9757.98866 4.017429e-01
## AMT_CREDIT 9807.46426 2.333029e-01
## AMT_ANNUITY 9841.48892 5.863881e-01
## NAME_TYPE_SUITEChildren 9835.74541 5.873021e-01
## NAME_TYPE_SUITEFamily 9851.82268 8.272514e-01
## NAME_TYPE_SUITEGroup of people 9854.89304 3.797799e-01
## NAME_TYPE_SUITEOther_A 9856.74596 3.913454e-01
## NAME_TYPE_SUITEOther_B 9853.70456 2.961447e-01
## NAME_TYPE_SUITESpouse, partner 9854.28935 8.385749e-01
## NAME_TYPE_SUITEUnaccompanied 9846.19867 7.314568e-01
## NAME_INCOME_TYPECommercial associate 9845.81218 7.179130e-01
## NAME_INCOME_TYPEPensioner 9850.37841 6.888786e-01
## NAME_INCOME_TYPEState servant 9846.11085 7.087370e-01
## NAME_INCOME_TYPEStudent 9858.09436 8.032757e-01
## NAME_INCOME_TYPEUnemployed 9852.56370 6.620542e-01
## NAME_INCOME_TYPEWorking 9845.49705 7.703314e-01
## NAME_EDUCATION_TYPEHigher education 9857.85728 7.618661e-01
## NAME_EDUCATION_TYPEIncomplete higher 9857.92865 6.587247e-01
## NAME_EDUCATION_TYPELower secondary 9856.82413 5.337013e-01
## NAME_EDUCATION_TYPESecondary / secondary special 9858.02816 6.491447e-01
## NAME_FAMILY_STATUSMarried 9851.19907 5.174350e-01
## NAME_FAMILY_STATUSSeparated 9852.71466 9.217954e-01
## NAME_FAMILY_STATUSSingle / not married 9855.64716 6.341287e-01
## NAME_FAMILY_STATUSWidow 9801.82300 3.956945e-01
## NAME_HOUSING_TYPEHouse / apartment 9850.15576 8.857889e-01
## NAME_HOUSING_TYPEMunicipal apartment 9852.04325 9.623446e-01
## NAME_HOUSING_TYPEOffice apartment 9855.89066 3.793102e-01
## NAME_HOUSING_TYPERented apartment 9857.25008 4.464711e-01
## NAME_HOUSING_TYPEWith parents 9853.18499 5.547401e-01
## REGION_POPULATION_RELATIVE 9810.09843 6.402163e-01
## DAYS_BIRTH 9852.36184 9.356256e-04
## DAYS_EMPLOYED 9851.66253 2.404757e-05
## OWN_CAR_AGE 62.44924 7.112027e-01
## OCCUPATION_TYPEAccountants 9795.14360 4.106112e-01
## OCCUPATION_TYPECleaning staff 9849.07232 4.875537e-01
## OCCUPATION_TYPECooking staff 9859.54353 4.998529e-01
## OCCUPATION_TYPECore staff 9857.26052 6.832844e-01
## OCCUPATION_TYPEDrivers 9849.62182 4.768261e-01
## OCCUPATION_TYPEHigh skill tech staff 9857.56269 2.553696e-01
## OCCUPATION_TYPEHR staff 9845.41388 9.177116e-01
## OCCUPATION_TYPEIT staff 9852.56666 6.420281e-01
## OCCUPATION_TYPELaborers 9838.74883 6.432848e-01
## OCCUPATION_TYPELow-skill Laborers 9839.07310 4.451489e-01
## OCCUPATION_TYPEManagers 9854.13464 6.724915e-01
## OCCUPATION_TYPEMedicine staff 9845.39368 2.763363e-01
## OCCUPATION_TYPEPrivate service staff 9857.24427 8.341408e-01
## OCCUPATION_TYPERealty agents 9858.84639 2.054845e-01
## OCCUPATION_TYPESales staff 9859.23310 7.405221e-01
## OCCUPATION_TYPESecretaries 9855.14485 3.925428e-01
## OCCUPATION_TYPESecurity staff 9850.20513 6.793137e-02
## OCCUPATION_TYPEWaiters/barmen staff 9857.58809 3.360059e-02
## CNT_FAM_MEMBERS2 9857.49802 9.383674e-01
## CNT_FAM_MEMBERS3 9858.30430 8.874652e-01
## CNT_FAM_MEMBERS4 9859.79692 7.580877e-01
## CNT_FAM_MEMBERS5 9859.60681 2.783340e-01
## CNT_FAM_MEMBERS6 9859.79036 3.301638e-02
## CNT_FAM_MEMBERS7 9859.90649 4.716909e-01
## ORGANIZATION_TYPEAgriculture 9851.96946 9.508535e-01
## ORGANIZATION_TYPEBank 9853.09656 9.071193e-01
## ORGANIZATION_TYPEBusiness Entity Type 1 9849.95852 9.856084e-01
## ORGANIZATION_TYPEBusiness Entity Type 2 9852.76155 5.787668e-01
## ORGANIZATION_TYPEBusiness Entity Type 3 9849.04393 8.899045e-01
## ORGANIZATION_TYPECleaning 9852.78481 7.417584e-01
## ORGANIZATION_TYPEConstruction 9846.97936 7.505159e-01
## ORGANIZATION_TYPECulture 9855.28352 5.577861e-01
## ORGANIZATION_TYPEElectricity 9855.34649 8.460544e-01
## ORGANIZATION_TYPEEmergency 9854.58245 8.132256e-01
## ORGANIZATION_TYPEGovernment 9849.30070 9.859606e-01
## ORGANIZATION_TYPEHotel 9856.82784 4.422898e-01
## ORGANIZATION_TYPEHousing 9853.55464 9.377354e-01
## ORGANIZATION_TYPEIndustry: type 1 9849.99755 6.318940e-01
## ORGANIZATION_TYPEIndustry: type 11 9848.57262 8.262021e-01
## ORGANIZATION_TYPEIndustry: type 12 9856.01441 3.293890e-01
## ORGANIZATION_TYPEIndustry: type 13 9852.40309 5.538987e-01
## ORGANIZATION_TYPEIndustry: type 2 9856.14096 4.181231e-01
## ORGANIZATION_TYPEIndustry: type 3 9849.73606 5.466176e-01
## ORGANIZATION_TYPEIndustry: type 4 9852.96031 8.674867e-01
## ORGANIZATION_TYPEIndustry: type 5 9855.15334 8.833089e-01
## ORGANIZATION_TYPEIndustry: type 6 9857.90329 5.717985e-01
## ORGANIZATION_TYPEIndustry: type 7 9848.23792 7.408390e-01
## ORGANIZATION_TYPEIndustry: type 9 9851.75824 6.530812e-01
## ORGANIZATION_TYPEInsurance 9853.96224 4.209860e-01
## ORGANIZATION_TYPEKindergarten 9850.80883 9.568605e-01
## ORGANIZATION_TYPELegal Services 9853.80085 8.391110e-01
## ORGANIZATION_TYPEMedicine 9851.10286 5.806277e-01
## ORGANIZATION_TYPEMilitary 9853.58736 6.072027e-01
## ORGANIZATION_TYPEMobile 9857.00555 3.816161e-01
## ORGANIZATION_TYPEOther 9850.35832 9.861743e-01
## ORGANIZATION_TYPEPolice 9852.53848 6.023603e-01
## ORGANIZATION_TYPEPostal 9845.47055 9.307302e-01
## ORGANIZATION_TYPERealtor 9857.75687 3.803680e-02
## ORGANIZATION_TYPEReligion 9858.83227 8.712232e-01
## ORGANIZATION_TYPERestaurant 9849.14060 8.105382e-01
## ORGANIZATION_TYPESchool 9850.11849 9.894174e-01
## ORGANIZATION_TYPESecurity 9851.16020 6.666244e-01
## ORGANIZATION_TYPESecurity Ministries 9849.88328 9.672298e-01
## ORGANIZATION_TYPESelf-employed 9850.71347 7.614641e-01
## ORGANIZATION_TYPEServices 9855.21952 8.871304e-01
## ORGANIZATION_TYPETelecom 9855.53402 8.184065e-01
## ORGANIZATION_TYPETrade: type 1 9847.12069 8.624781e-01
## ORGANIZATION_TYPETrade: type 2 9850.05884 9.883329e-01
## ORGANIZATION_TYPETrade: type 3 9842.19071 8.246838e-01
## ORGANIZATION_TYPETrade: type 4 9858.41253 6.191633e-01
## ORGANIZATION_TYPETrade: type 6 9850.30181 9.156251e-01
## ORGANIZATION_TYPETrade: type 7 9848.48774 5.766373e-01
## ORGANIZATION_TYPETransport: type 1 9857.65978 6.193872e-01
## ORGANIZATION_TYPETransport: type 2 9852.87199 2.766958e-01
## ORGANIZATION_TYPETransport: type 3 9854.21526 7.104211e-02
## ORGANIZATION_TYPETransport: type 4 9849.40373 6.128549e-01
## ORGANIZATION_TYPEUniversity 9851.63858 9.962011e-01
## ORGANIZATION_TYPEXNA 9855.66774 2.347956e-04
## AMR_REQ_CREDIT_BUREAU_SUM1 1417.16708 6.461736e-01
## AMR_REQ_CREDIT_BUREAU_SUM2 250.28318 2.658015e-01
## AMR_REQ_CREDIT_BUREAU_SUM3 260.84440 2.211437e-01
## AMR_REQ_CREDIT_BUREAU_SUM4 311.38428 7.881814e-01
## AMR_REQ_CREDIT_BUREAU_SUM5 499.95796 6.063039e-01
## AMR_REQ_CREDIT_BUREAU_SUM6 172.24968 1.468333e-01
## AMR_REQ_CREDIT_BUREAU_SUM7 521.47606 4.028441e-01
## AMR_REQ_CREDIT_BUREAU_SUM8 302.47679 3.295141e-02
## AMR_REQ_CREDIT_BUREAU_SUM9 276.96744 6.392286e-01
## AMR_REQ_CREDIT_BUREAU_SUM10 6861.17223 6.789695e-01
## AMR_REQ_CREDIT_BUREAU_SUM11 1787.91692 9.652245e-01
## AMR_REQ_CREDIT_BUREAU_SUM12 9659.66009 6.953321e-01
## AMR_REQ_CREDIT_BUREAU_SUM13 9710.15537 5.256008e-01
## AMR_REQ_CREDIT_BUREAU_SUM14 9819.14406 9.226638e-01
## AMR_REQ_CREDIT_BUREAU_SUM15 9501.72835 5.299607e-01
## AMR_REQ_CREDIT_BUREAU_SUM16 9859.90649 9.196110e-01
## AMR_REQ_CREDIT_BUREAU_SUM17 9859.35814 7.155472e-01
## AMR_REQ_CREDIT_BUREAU_SUM18 9491.65927 6.899178e-01
## AMR_REQ_CREDIT_BUREAU_SUM19 9859.90649 9.511950e-01
## AMR_REQ_CREDIT_BUREAU_SUM20 9858.65550 8.435467e-01
## AMR_REQ_CREDIT_BUREAU_SUM28 9833.27900 6.672155e-01summary(pool(reg), conf.int=T)## estimate
## (Intercept) 2.063259e-01
## CODE_GENDERM 4.352771e-02
## NAME_CONTRACT_TYPERevolving loans -5.055284e-02
## AMT_INCOME_TOTAL -5.315301e-08
## FLAG_OWN_CARY -3.081047e-02
## FLAG_OWN_REALTYY 5.206398e-03
## AMT_CREDIT -1.320492e-08
## AMT_ANNUITY -1.769943e-07
## NAME_TYPE_SUITEChildren -2.756176e-02
## NAME_TYPE_SUITEFamily -9.752861e-03
## NAME_TYPE_SUITEGroup of people -1.140243e-01
## NAME_TYPE_SUITEOther_A -6.362593e-02
## NAME_TYPE_SUITEOther_B -6.131117e-02
## NAME_TYPE_SUITESpouse, partner -9.428628e-03
## NAME_TYPE_SUITEUnaccompanied -1.514317e-02
## NAME_INCOME_TYPECommercial associate -1.027763e-01
## NAME_INCOME_TYPEPensioner -1.580174e-01
## NAME_INCOME_TYPEState servant -1.063998e-01
## NAME_INCOME_TYPEStudent -9.719079e-02
## NAME_INCOME_TYPEUnemployed -2.097608e-01
## NAME_INCOME_TYPEWorking -8.310605e-02
## NAME_EDUCATION_TYPEHigher education 3.705344e-02
## NAME_EDUCATION_TYPEIncomplete higher 5.441029e-02
## NAME_EDUCATION_TYPELower secondary 7.767959e-02
## NAME_EDUCATION_TYPESecondary / secondary special 5.559630e-02
## NAME_FAMILY_STATUSMarried -6.035944e-03
## NAME_FAMILY_STATUSSeparated -1.670707e-03
## NAME_FAMILY_STATUSSingle / not married 7.999912e-03
## NAME_FAMILY_STATUSWidow 1.727822e-02
## NAME_HOUSING_TYPEHouse / apartment 5.891942e-03
## NAME_HOUSING_TYPEMunicipal apartment -2.051035e-03
## NAME_HOUSING_TYPEOffice apartment -4.512994e-02
## NAME_HOUSING_TYPERented apartment 3.482144e-02
## NAME_HOUSING_TYPEWith parents 2.526804e-02
## REGION_POPULATION_RELATIVE -9.805655e-02
## DAYS_BIRTH 3.216099e-06
## DAYS_EMPLOYED 6.241027e-06
## OWN_CAR_AGE 1.037233e-04
## OCCUPATION_TYPEAccountants -1.463772e-02
## OCCUPATION_TYPECleaning staff -1.694824e-02
## OCCUPATION_TYPECooking staff 1.418259e-02
## OCCUPATION_TYPECore staff -5.402408e-03
## OCCUPATION_TYPEDrivers 1.028013e-02
## OCCUPATION_TYPEHigh skill tech staff -1.774165e-02
## OCCUPATION_TYPEHR staff -7.604784e-03
## OCCUPATION_TYPEIT staff -2.806577e-02
## OCCUPATION_TYPELaborers -4.816502e-03
## OCCUPATION_TYPELow-skill Laborers 2.681807e-02
## OCCUPATION_TYPEManagers -5.676823e-03
## OCCUPATION_TYPEMedicine staff -2.491503e-02
## OCCUPATION_TYPEPrivate service staff -6.379935e-03
## OCCUPATION_TYPERealty agents 7.190980e-02
## OCCUPATION_TYPESales staff -4.065929e-03
## OCCUPATION_TYPESecretaries 4.072782e-02
## OCCUPATION_TYPESecurity staff 4.084883e-02
## OCCUPATION_TYPEWaiters/barmen staff 1.002301e-01
## CNT_FAM_MEMBERS2 1.132116e-03
## CNT_FAM_MEMBERS3 -2.320473e-03
## CNT_FAM_MEMBERS4 -5.599006e-03
## CNT_FAM_MEMBERS5 3.401739e-02
## CNT_FAM_MEMBERS6 1.794479e-01
## CNT_FAM_MEMBERS7 8.906367e-02
## ORGANIZATION_TYPEAgriculture 4.382414e-03
## ORGANIZATION_TYPEBank 8.570364e-03
## ORGANIZATION_TYPEBusiness Entity Type 1 -1.214637e-03
## ORGANIZATION_TYPEBusiness Entity Type 2 3.669097e-02
## ORGANIZATION_TYPEBusiness Entity Type 3 8.953085e-03
## ORGANIZATION_TYPECleaning 3.671069e-02
## ORGANIZATION_TYPEConstruction 2.132647e-02
## ORGANIZATION_TYPECulture -5.861336e-02
## ORGANIZATION_TYPEElectricity -1.564123e-02
## ORGANIZATION_TYPEEmergency -2.077316e-02
## ORGANIZATION_TYPEGovernment 1.162231e-03
## ORGANIZATION_TYPEHotel -6.344843e-02
## ORGANIZATION_TYPEHousing 5.454882e-03
## ORGANIZATION_TYPEIndustry: type 1 -3.756790e-02
## ORGANIZATION_TYPEIndustry: type 11 1.558636e-02
## ORGANIZATION_TYPEIndustry: type 12 9.957355e-02
## ORGANIZATION_TYPEIndustry: type 13 -1.205705e-01
## ORGANIZATION_TYPEIndustry: type 2 -7.879773e-02
## ORGANIZATION_TYPEIndustry: type 3 4.168363e-02
## ORGANIZATION_TYPEIndustry: type 4 1.304441e-02
## ORGANIZATION_TYPEIndustry: type 5 1.303146e-02
## ORGANIZATION_TYPEIndustry: type 6 -7.803666e-02
## ORGANIZATION_TYPEIndustry: type 7 2.541053e-02
## ORGANIZATION_TYPEIndustry: type 9 3.122153e-02
## ORGANIZATION_TYPEInsurance -7.338810e-02
## ORGANIZATION_TYPEKindergarten -3.629599e-03
## ORGANIZATION_TYPELegal Services 1.982034e-02
## ORGANIZATION_TYPEMedicine 3.709787e-02
## ORGANIZATION_TYPEMilitary -3.655942e-02
## ORGANIZATION_TYPEMobile -9.156294e-02
## ORGANIZATION_TYPEOther 1.135235e-03
## ORGANIZATION_TYPEPolice -3.770840e-02
## ORGANIZATION_TYPEPostal -6.280312e-03
## ORGANIZATION_TYPERealtor 2.169619e-01
## ORGANIZATION_TYPEReligion -3.296117e-02
## ORGANIZATION_TYPERestaurant -1.767501e-02
## ORGANIZATION_TYPESchool 8.833369e-04
## ORGANIZATION_TYPESecurity -3.087008e-02
## ORGANIZATION_TYPESecurity Ministries -3.000705e-03
## ORGANIZATION_TYPESelf-employed 1.969723e-02
## ORGANIZATION_TYPEServices -1.091431e-02
## ORGANIZATION_TYPETelecom -2.065610e-02
## ORGANIZATION_TYPETrade: type 1 1.810593e-02
## ORGANIZATION_TYPETrade: type 2 -1.085085e-03
## ORGANIZATION_TYPETrade: type 3 -1.529725e-02
## ORGANIZATION_TYPETrade: type 4 -1.012080e-01
## ORGANIZATION_TYPETrade: type 6 -9.123168e-03
## ORGANIZATION_TYPETrade: type 7 3.726542e-02
## ORGANIZATION_TYPETransport: type 1 -7.505025e-02
## ORGANIZATION_TYPETransport: type 2 7.819173e-02
## ORGANIZATION_TYPETransport: type 3 1.384465e-01
## ORGANIZATION_TYPETransport: type 4 -3.459733e-02
## ORGANIZATION_TYPEUniversity -3.775120e-04
## ORGANIZATION_TYPEXNA -2.240731e+00
## AMR_REQ_CREDIT_BUREAU_SUM1 4.190379e-03
## AMR_REQ_CREDIT_BUREAU_SUM2 1.092473e-02
## AMR_REQ_CREDIT_BUREAU_SUM3 1.288864e-02
## AMR_REQ_CREDIT_BUREAU_SUM4 3.119740e-03
## AMR_REQ_CREDIT_BUREAU_SUM5 -7.012164e-03
## AMR_REQ_CREDIT_BUREAU_SUM6 2.503563e-02
## AMR_REQ_CREDIT_BUREAU_SUM7 1.588082e-02
## AMR_REQ_CREDIT_BUREAU_SUM8 5.843887e-02
## AMR_REQ_CREDIT_BUREAU_SUM9 1.691654e-02
## AMR_REQ_CREDIT_BUREAU_SUM10 -1.849285e-02
## AMR_REQ_CREDIT_BUREAU_SUM11 3.053713e-03
## AMR_REQ_CREDIT_BUREAU_SUM12 -4.140642e-02
## AMR_REQ_CREDIT_BUREAU_SUM13 -7.344946e-02
## AMR_REQ_CREDIT_BUREAU_SUM14 -1.746657e-02
## AMR_REQ_CREDIT_BUREAU_SUM15 -8.302970e-02
## AMR_REQ_CREDIT_BUREAU_SUM16 -2.768300e-02
## AMR_REQ_CREDIT_BUREAU_SUM17 -9.937823e-02
## AMR_REQ_CREDIT_BUREAU_SUM18 -7.341512e-02
## AMR_REQ_CREDIT_BUREAU_SUM19 -1.680945e-02
## AMR_REQ_CREDIT_BUREAU_SUM20 -5.056997e-02
## AMR_REQ_CREDIT_BUREAU_SUM28 -1.117664e-01
## std.error statistic
## (Intercept) 3.238824e-01 0.637039695
## CODE_GENDERM 7.349921e-03 5.922200671
## NAME_CONTRACT_TYPERevolving loans 9.804538e-03 -5.156065232
## AMT_INCOME_TOTAL 3.806233e-08 -1.396472747
## FLAG_OWN_CARY 6.335356e-03 -4.863258282
## FLAG_OWN_REALTYY 6.208820e-03 0.838548695
## AMT_CREDIT 1.107826e-08 -1.191966886
## AMT_ANNUITY 3.253000e-07 -0.544095497
## NAME_TYPE_SUITEChildren 5.078004e-02 -0.542767609
## NAME_TYPE_SUITEFamily 4.468996e-02 -0.218233848
## NAME_TYPE_SUITEGroup of people 1.298177e-01 -0.878341280
## NAME_TYPE_SUITEOther_A 7.422377e-02 -0.857217725
## NAME_TYPE_SUITEOther_B 5.868265e-02 -1.044792141
## NAME_TYPE_SUITESpouse, partner 4.628181e-02 -0.203722120
## NAME_TYPE_SUITEUnaccompanied 4.412372e-02 -0.343197977
## NAME_INCOME_TYPECommercial associate 2.844942e-01 -0.361259798
## NAME_INCOME_TYPEPensioner 3.946596e-01 -0.400389083
## NAME_INCOME_TYPEState servant 2.848237e-01 -0.373563697
## NAME_INCOME_TYPEStudent 3.901416e-01 -0.249116717
## NAME_INCOME_TYPEUnemployed 4.799008e-01 -0.437092064
## NAME_INCOME_TYPEWorking 2.846590e-01 -0.291949486
## NAME_EDUCATION_TYPEHigher education 1.222726e-01 0.303039512
## NAME_EDUCATION_TYPEIncomplete higher 1.231872e-01 0.441687961
## NAME_EDUCATION_TYPELower secondary 1.248090e-01 0.622387764
## NAME_EDUCATION_TYPESecondary / secondary special 1.221992e-01 0.454964697
## NAME_FAMILY_STATUSMarried 9.324405e-03 -0.647327578
## NAME_FAMILY_STATUSSeparated 1.701766e-02 -0.098174855
## NAME_FAMILY_STATUSSingle / not married 1.680871e-02 0.475938598
## NAME_FAMILY_STATUSWidow 2.034232e-02 0.849372761
## NAME_HOUSING_TYPEHouse / apartment 4.101925e-02 0.143638448
## NAME_HOUSING_TYPEMunicipal apartment 4.344237e-02 -0.047212774
## NAME_HOUSING_TYPEOffice apartment 5.133026e-02 -0.879207432
## NAME_HOUSING_TYPERented apartment 4.573693e-02 0.761341852
## NAME_HOUSING_TYPEWith parents 4.277699e-02 0.590692396
## REGION_POPULATION_RELATIVE 2.097866e-01 -0.467410822
## DAYS_BIRTH 9.715712e-07 3.310204273
## DAYS_EMPLOYED 1.476965e-06 4.225575963
## OWN_CAR_AGE 2.801423e-04 0.370252048
## OCCUPATION_TYPEAccountants 1.778898e-02 -0.822853488
## OCCUPATION_TYPECleaning staff 2.441296e-02 -0.694231084
## OCCUPATION_TYPECooking staff 2.101915e-02 0.674746192
## OCCUPATION_TYPECore staff 1.324128e-02 -0.407997419
## OCCUPATION_TYPEDrivers 1.444968e-02 0.711443845
## OCCUPATION_TYPEHigh skill tech staff 1.559747e-02 -1.137469734
## OCCUPATION_TYPEHR staff 7.360465e-02 -0.103319349
## OCCUPATION_TYPEIT staff 6.037217e-02 -0.464879335
## OCCUPATION_TYPELaborers 1.040000e-02 -0.463125263
## OCCUPATION_TYPELow-skill Laborers 3.512250e-02 0.763558124
## OCCUPATION_TYPEManagers 1.342852e-02 -0.422743750
## OCCUPATION_TYPEMedicine staff 2.288624e-02 -1.088646858
## OCCUPATION_TYPEPrivate service staff 3.046778e-02 -0.209399389
## OCCUPATION_TYPERealty agents 5.679346e-02 1.266163399
## OCCUPATION_TYPESales staff 1.227742e-02 -0.331171413
## OCCUPATION_TYPESecretaries 4.763194e-02 0.855052630
## OCCUPATION_TYPESecurity staff 2.237480e-02 1.825662323
## OCCUPATION_TYPEWaiters/barmen staff 4.716424e-02 2.125129651
## CNT_FAM_MEMBERS2 1.464121e-02 0.077323907
## CNT_FAM_MEMBERS3 1.639724e-02 -0.141516061
## CNT_FAM_MEMBERS4 1.817850e-02 -0.308001501
## CNT_FAM_MEMBERS5 3.137761e-02 1.084129310
## CNT_FAM_MEMBERS6 8.416156e-02 2.132183844
## CNT_FAM_MEMBERS7 1.237411e-01 0.719758220
## ORGANIZATION_TYPEAgriculture 7.110094e-02 0.061636519
## ORGANIZATION_TYPEBank 7.345444e-02 0.116675911
## ORGANIZATION_TYPEBusiness Entity Type 1 6.733533e-02 -0.018038620
## ORGANIZATION_TYPEBusiness Entity Type 2 6.608540e-02 0.555205375
## ORGANIZATION_TYPEBusiness Entity Type 3 6.467652e-02 0.138428667
## ORGANIZATION_TYPECleaning 1.114015e-01 0.329534866
## ORGANIZATION_TYPEConstruction 6.707111e-02 0.317968028
## ORGANIZATION_TYPECulture 9.999669e-02 -0.586153024
## ORGANIZATION_TYPEElectricity 8.055834e-02 -0.194160226
## ORGANIZATION_TYPEEmergency 8.792009e-02 -0.236273141
## ORGANIZATION_TYPEGovernment 6.604658e-02 0.017597149
## ORGANIZATION_TYPEHotel 8.257608e-02 -0.768363320
## ORGANIZATION_TYPEHousing 6.982835e-02 0.078118440
## ORGANIZATION_TYPEIndustry: type 1 7.841714e-02 -0.479077683
## ORGANIZATION_TYPEIndustry: type 11 7.098229e-02 0.219580928
## ORGANIZATION_TYPEIndustry: type 12 1.020855e-01 0.975393521
## ORGANIZATION_TYPEIndustry: type 13 2.036842e-01 -0.591948495
## ORGANIZATION_TYPEIndustry: type 2 9.731529e-02 -0.809715791
## ORGANIZATION_TYPEIndustry: type 3 6.914345e-02 0.602857174
## ORGANIZATION_TYPEIndustry: type 4 7.817763e-02 0.166856096
## ORGANIZATION_TYPEIndustry: type 5 8.878250e-02 0.146779649
## ORGANIZATION_TYPEIndustry: type 6 1.380145e-01 -0.565423688
## ORGANIZATION_TYPEIndustry: type 7 7.682659e-02 0.330751780
## ORGANIZATION_TYPEIndustry: type 9 6.945843e-02 0.449499499
## ORGANIZATION_TYPEInsurance 9.119419e-02 -0.804745374
## ORGANIZATION_TYPEKindergarten 6.709661e-02 -0.054095110
## ORGANIZATION_TYPELegal Services 9.761973e-02 0.203036166
## ORGANIZATION_TYPEMedicine 6.714712e-02 0.552486377
## ORGANIZATION_TYPEMilitary 7.111523e-02 -0.514086997
## ORGANIZATION_TYPEMobile 1.046480e-01 -0.874961544
## ORGANIZATION_TYPEOther 6.550997e-02 0.017329199
## ORGANIZATION_TYPEPolice 7.237334e-02 -0.521026129
## ORGANIZATION_TYPEPostal 7.224699e-02 -0.086928360
## ORGANIZATION_TYPERealtor 1.045732e-01 2.074736910
## ORGANIZATION_TYPEReligion 2.033270e-01 -0.162109178
## ORGANIZATION_TYPERestaurant 7.372632e-02 -0.239738171
## ORGANIZATION_TYPESchool 6.659657e-02 0.013263999
## ORGANIZATION_TYPESecurity 7.165784e-02 -0.430798349
## ORGANIZATION_TYPESecurity Ministries 7.303849e-02 -0.041083890
## ORGANIZATION_TYPESelf-employed 6.488590e-02 0.303567190
## ORGANIZATION_TYPEServices 7.689387e-02 -0.141939911
## ORGANIZATION_TYPETelecom 8.996517e-02 -0.229601124
## ORGANIZATION_TYPETrade: type 1 1.045225e-01 0.173225100
## ORGANIZATION_TYPETrade: type 2 7.420166e-02 -0.014623457
## ORGANIZATION_TYPETrade: type 3 6.905253e-02 -0.221530686
## ORGANIZATION_TYPETrade: type 4 2.036165e-01 -0.497052183
## ORGANIZATION_TYPETrade: type 6 8.610919e-02 -0.105948836
## ORGANIZATION_TYPETrade: type 7 6.674541e-02 0.558321819
## ORGANIZATION_TYPETransport: type 1 1.510872e-01 -0.496734785
## ORGANIZATION_TYPETransport: type 2 7.187849e-02 1.087832087
## ORGANIZATION_TYPETransport: type 3 7.668451e-02 1.805403139
## ORGANIZATION_TYPETransport: type 4 6.837164e-02 -0.506018664
## ORGANIZATION_TYPEUniversity 7.928623e-02 -0.004761382
## ORGANIZATION_TYPEXNA 6.089517e-01 -3.679653516
## AMR_REQ_CREDIT_BUREAU_SUM1 9.127409e-03 0.459098434
## AMR_REQ_CREDIT_BUREAU_SUM2 9.816924e-03 1.112846530
## AMR_REQ_CREDIT_BUREAU_SUM3 1.053364e-02 1.223569960
## AMR_REQ_CREDIT_BUREAU_SUM4 1.161135e-02 0.268680246
## AMR_REQ_CREDIT_BUREAU_SUM5 1.360600e-02 -0.515373121
## AMR_REQ_CREDIT_BUREAU_SUM6 1.725497e-02 1.450922280
## AMR_REQ_CREDIT_BUREAU_SUM7 1.898283e-02 0.836588629
## AMR_REQ_CREDIT_BUREAU_SUM8 2.739782e-02 2.132974918
## AMR_REQ_CREDIT_BUREAU_SUM9 3.608539e-02 0.468792048
## AMR_REQ_CREDIT_BUREAU_SUM10 4.468145e-02 -0.413882025
## AMR_REQ_CREDIT_BUREAU_SUM11 7.004005e-02 0.043599527
## AMR_REQ_CREDIT_BUREAU_SUM12 1.057255e-01 -0.391640785
## AMR_REQ_CREDIT_BUREAU_SUM13 1.157125e-01 -0.634758237
## AMR_REQ_CREDIT_BUREAU_SUM14 1.799169e-01 -0.097081323
## AMR_REQ_CREDIT_BUREAU_SUM15 1.321943e-01 -0.628088210
## AMR_REQ_CREDIT_BUREAU_SUM16 2.742893e-01 -0.100926295
## AMR_REQ_CREDIT_BUREAU_SUM17 2.726974e-01 -0.364426748
## AMR_REQ_CREDIT_BUREAU_SUM18 1.840078e-01 -0.398978291
## AMR_REQ_CREDIT_BUREAU_SUM19 2.746293e-01 -0.061207775
## AMR_REQ_CREDIT_BUREAU_SUM20 2.562267e-01 -0.197364150
## AMR_REQ_CREDIT_BUREAU_SUM28 2.599306e-01 -0.429985680
## df p.value
## (Intercept) 9842.13067 5.241138e-01
## CODE_GENDERM 9847.07017 3.282767e-09
## NAME_CONTRACT_TYPERevolving loans 9667.51517 2.570858e-07
## AMT_INCOME_TOTAL 9030.00715 1.626036e-01
## FLAG_OWN_CARY 9859.54073 1.172539e-06
## FLAG_OWN_REALTYY 9757.98866 4.017429e-01
## AMT_CREDIT 9807.46426 2.333029e-01
## AMT_ANNUITY 9841.48892 5.863881e-01
## NAME_TYPE_SUITEChildren 9835.74541 5.873021e-01
## NAME_TYPE_SUITEFamily 9851.82268 8.272514e-01
## NAME_TYPE_SUITEGroup of people 9854.89304 3.797799e-01
## NAME_TYPE_SUITEOther_A 9856.74596 3.913454e-01
## NAME_TYPE_SUITEOther_B 9853.70456 2.961447e-01
## NAME_TYPE_SUITESpouse, partner 9854.28935 8.385749e-01
## NAME_TYPE_SUITEUnaccompanied 9846.19867 7.314568e-01
## NAME_INCOME_TYPECommercial associate 9845.81218 7.179130e-01
## NAME_INCOME_TYPEPensioner 9850.37841 6.888786e-01
## NAME_INCOME_TYPEState servant 9846.11085 7.087370e-01
## NAME_INCOME_TYPEStudent 9858.09436 8.032757e-01
## NAME_INCOME_TYPEUnemployed 9852.56370 6.620542e-01
## NAME_INCOME_TYPEWorking 9845.49705 7.703314e-01
## NAME_EDUCATION_TYPEHigher education 9857.85728 7.618661e-01
## NAME_EDUCATION_TYPEIncomplete higher 9857.92865 6.587247e-01
## NAME_EDUCATION_TYPELower secondary 9856.82413 5.337013e-01
## NAME_EDUCATION_TYPESecondary / secondary special 9858.02816 6.491447e-01
## NAME_FAMILY_STATUSMarried 9851.19907 5.174350e-01
## NAME_FAMILY_STATUSSeparated 9852.71466 9.217954e-01
## NAME_FAMILY_STATUSSingle / not married 9855.64716 6.341287e-01
## NAME_FAMILY_STATUSWidow 9801.82300 3.956945e-01
## NAME_HOUSING_TYPEHouse / apartment 9850.15576 8.857889e-01
## NAME_HOUSING_TYPEMunicipal apartment 9852.04325 9.623446e-01
## NAME_HOUSING_TYPEOffice apartment 9855.89066 3.793102e-01
## NAME_HOUSING_TYPERented apartment 9857.25008 4.464711e-01
## NAME_HOUSING_TYPEWith parents 9853.18499 5.547401e-01
## REGION_POPULATION_RELATIVE 9810.09843 6.402163e-01
## DAYS_BIRTH 9852.36184 9.356256e-04
## DAYS_EMPLOYED 9851.66253 2.404757e-05
## OWN_CAR_AGE 62.44924 7.112027e-01
## OCCUPATION_TYPEAccountants 9795.14360 4.106112e-01
## OCCUPATION_TYPECleaning staff 9849.07232 4.875537e-01
## OCCUPATION_TYPECooking staff 9859.54353 4.998529e-01
## OCCUPATION_TYPECore staff 9857.26052 6.832844e-01
## OCCUPATION_TYPEDrivers 9849.62182 4.768261e-01
## OCCUPATION_TYPEHigh skill tech staff 9857.56269 2.553696e-01
## OCCUPATION_TYPEHR staff 9845.41388 9.177116e-01
## OCCUPATION_TYPEIT staff 9852.56666 6.420281e-01
## OCCUPATION_TYPELaborers 9838.74883 6.432848e-01
## OCCUPATION_TYPELow-skill Laborers 9839.07310 4.451489e-01
## OCCUPATION_TYPEManagers 9854.13464 6.724915e-01
## OCCUPATION_TYPEMedicine staff 9845.39368 2.763363e-01
## OCCUPATION_TYPEPrivate service staff 9857.24427 8.341408e-01
## OCCUPATION_TYPERealty agents 9858.84639 2.054845e-01
## OCCUPATION_TYPESales staff 9859.23310 7.405221e-01
## OCCUPATION_TYPESecretaries 9855.14485 3.925428e-01
## OCCUPATION_TYPESecurity staff 9850.20513 6.793137e-02
## OCCUPATION_TYPEWaiters/barmen staff 9857.58809 3.360059e-02
## CNT_FAM_MEMBERS2 9857.49802 9.383674e-01
## CNT_FAM_MEMBERS3 9858.30430 8.874652e-01
## CNT_FAM_MEMBERS4 9859.79692 7.580877e-01
## CNT_FAM_MEMBERS5 9859.60681 2.783340e-01
## CNT_FAM_MEMBERS6 9859.79036 3.301638e-02
## CNT_FAM_MEMBERS7 9859.90649 4.716909e-01
## ORGANIZATION_TYPEAgriculture 9851.96946 9.508535e-01
## ORGANIZATION_TYPEBank 9853.09656 9.071193e-01
## ORGANIZATION_TYPEBusiness Entity Type 1 9849.95852 9.856084e-01
## ORGANIZATION_TYPEBusiness Entity Type 2 9852.76155 5.787668e-01
## ORGANIZATION_TYPEBusiness Entity Type 3 9849.04393 8.899045e-01
## ORGANIZATION_TYPECleaning 9852.78481 7.417584e-01
## ORGANIZATION_TYPEConstruction 9846.97936 7.505159e-01
## ORGANIZATION_TYPECulture 9855.28352 5.577861e-01
## ORGANIZATION_TYPEElectricity 9855.34649 8.460544e-01
## ORGANIZATION_TYPEEmergency 9854.58245 8.132256e-01
## ORGANIZATION_TYPEGovernment 9849.30070 9.859606e-01
## ORGANIZATION_TYPEHotel 9856.82784 4.422898e-01
## ORGANIZATION_TYPEHousing 9853.55464 9.377354e-01
## ORGANIZATION_TYPEIndustry: type 1 9849.99755 6.318940e-01
## ORGANIZATION_TYPEIndustry: type 11 9848.57262 8.262021e-01
## ORGANIZATION_TYPEIndustry: type 12 9856.01441 3.293890e-01
## ORGANIZATION_TYPEIndustry: type 13 9852.40309 5.538987e-01
## ORGANIZATION_TYPEIndustry: type 2 9856.14096 4.181231e-01
## ORGANIZATION_TYPEIndustry: type 3 9849.73606 5.466176e-01
## ORGANIZATION_TYPEIndustry: type 4 9852.96031 8.674867e-01
## ORGANIZATION_TYPEIndustry: type 5 9855.15334 8.833089e-01
## ORGANIZATION_TYPEIndustry: type 6 9857.90329 5.717985e-01
## ORGANIZATION_TYPEIndustry: type 7 9848.23792 7.408390e-01
## ORGANIZATION_TYPEIndustry: type 9 9851.75824 6.530812e-01
## ORGANIZATION_TYPEInsurance 9853.96224 4.209860e-01
## ORGANIZATION_TYPEKindergarten 9850.80883 9.568605e-01
## ORGANIZATION_TYPELegal Services 9853.80085 8.391110e-01
## ORGANIZATION_TYPEMedicine 9851.10286 5.806277e-01
## ORGANIZATION_TYPEMilitary 9853.58736 6.072027e-01
## ORGANIZATION_TYPEMobile 9857.00555 3.816161e-01
## ORGANIZATION_TYPEOther 9850.35832 9.861743e-01
## ORGANIZATION_TYPEPolice 9852.53848 6.023603e-01
## ORGANIZATION_TYPEPostal 9845.47055 9.307302e-01
## ORGANIZATION_TYPERealtor 9857.75687 3.803680e-02
## ORGANIZATION_TYPEReligion 9858.83227 8.712232e-01
## ORGANIZATION_TYPERestaurant 9849.14060 8.105382e-01
## ORGANIZATION_TYPESchool 9850.11849 9.894174e-01
## ORGANIZATION_TYPESecurity 9851.16020 6.666244e-01
## ORGANIZATION_TYPESecurity Ministries 9849.88328 9.672298e-01
## ORGANIZATION_TYPESelf-employed 9850.71347 7.614641e-01
## ORGANIZATION_TYPEServices 9855.21952 8.871304e-01
## ORGANIZATION_TYPETelecom 9855.53402 8.184065e-01
## ORGANIZATION_TYPETrade: type 1 9847.12069 8.624781e-01
## ORGANIZATION_TYPETrade: type 2 9850.05884 9.883329e-01
## ORGANIZATION_TYPETrade: type 3 9842.19071 8.246838e-01
## ORGANIZATION_TYPETrade: type 4 9858.41253 6.191633e-01
## ORGANIZATION_TYPETrade: type 6 9850.30181 9.156251e-01
## ORGANIZATION_TYPETrade: type 7 9848.48774 5.766373e-01
## ORGANIZATION_TYPETransport: type 1 9857.65978 6.193872e-01
## ORGANIZATION_TYPETransport: type 2 9852.87199 2.766958e-01
## ORGANIZATION_TYPETransport: type 3 9854.21526 7.104211e-02
## ORGANIZATION_TYPETransport: type 4 9849.40373 6.128549e-01
## ORGANIZATION_TYPEUniversity 9851.63858 9.962011e-01
## ORGANIZATION_TYPEXNA 9855.66774 2.347956e-04
## AMR_REQ_CREDIT_BUREAU_SUM1 1417.16708 6.461736e-01
## AMR_REQ_CREDIT_BUREAU_SUM2 250.28318 2.658015e-01
## AMR_REQ_CREDIT_BUREAU_SUM3 260.84440 2.211437e-01
## AMR_REQ_CREDIT_BUREAU_SUM4 311.38428 7.881814e-01
## AMR_REQ_CREDIT_BUREAU_SUM5 499.95796 6.063039e-01
## AMR_REQ_CREDIT_BUREAU_SUM6 172.24968 1.468333e-01
## AMR_REQ_CREDIT_BUREAU_SUM7 521.47606 4.028441e-01
## AMR_REQ_CREDIT_BUREAU_SUM8 302.47679 3.295141e-02
## AMR_REQ_CREDIT_BUREAU_SUM9 276.96744 6.392286e-01
## AMR_REQ_CREDIT_BUREAU_SUM10 6861.17223 6.789695e-01
## AMR_REQ_CREDIT_BUREAU_SUM11 1787.91692 9.652245e-01
## AMR_REQ_CREDIT_BUREAU_SUM12 9659.66009 6.953321e-01
## AMR_REQ_CREDIT_BUREAU_SUM13 9710.15537 5.256008e-01
## AMR_REQ_CREDIT_BUREAU_SUM14 9819.14406 9.226638e-01
## AMR_REQ_CREDIT_BUREAU_SUM15 9501.72835 5.299607e-01
## AMR_REQ_CREDIT_BUREAU_SUM16 9859.90649 9.196110e-01
## AMR_REQ_CREDIT_BUREAU_SUM17 9859.35814 7.155472e-01
## AMR_REQ_CREDIT_BUREAU_SUM18 9491.65927 6.899178e-01
## AMR_REQ_CREDIT_BUREAU_SUM19 9859.90649 9.511950e-01
## AMR_REQ_CREDIT_BUREAU_SUM20 9858.65550 8.435467e-01
## AMR_REQ_CREDIT_BUREAU_SUM28 9833.27900 6.672155e-01
## 2.5 %
## (Intercept) -4.285499e-01
## CODE_GENDERM 2.912035e-02
## NAME_CONTRACT_TYPERevolving loans -6.977179e-02
## AMT_INCOME_TOTAL -1.277638e-07
## FLAG_OWN_CARY -4.322907e-02
## FLAG_OWN_REALTYY -6.964175e-03
## AMT_CREDIT -3.492059e-08
## AMT_ANNUITY -8.146490e-07
## NAME_TYPE_SUITEChildren -1.271011e-01
## NAME_TYPE_SUITEFamily -9.735433e-02
## NAME_TYPE_SUITEGroup of people -3.684936e-01
## NAME_TYPE_SUITEOther_A -2.091197e-01
## NAME_TYPE_SUITEOther_B -1.763412e-01
## NAME_TYPE_SUITESpouse, partner -1.001504e-01
## NAME_TYPE_SUITEUnaccompanied -1.016347e-01
## NAME_INCOME_TYPECommercial associate -6.604432e-01
## NAME_INCOME_TYPEPensioner -9.316310e-01
## NAME_INCOME_TYPEState servant -6.647126e-01
## NAME_INCOME_TYPEStudent -8.619482e-01
## NAME_INCOME_TYPEUnemployed -1.150465e+00
## NAME_INCOME_TYPEWorking -6.410961e-01
## NAME_EDUCATION_TYPEHigher education -2.026260e-01
## NAME_EDUCATION_TYPEIncomplete higher -1.870618e-01
## NAME_EDUCATION_TYPELower secondary -1.669716e-01
## NAME_EDUCATION_TYPESecondary / secondary special -1.839391e-01
## NAME_FAMILY_STATUSMarried -2.431369e-02
## NAME_FAMILY_STATUSSeparated -3.502881e-02
## NAME_FAMILY_STATUSSingle / not married -2.494859e-02
## NAME_FAMILY_STATUSWidow -2.259693e-02
## NAME_HOUSING_TYPEHouse / apartment -7.451419e-02
## NAME_HOUSING_TYPEMunicipal apartment -8.720697e-02
## NAME_HOUSING_TYPEOffice apartment -1.457478e-01
## NAME_HOUSING_TYPERented apartment -5.483231e-02
## NAME_HOUSING_TYPEWith parents -5.858361e-02
## REGION_POPULATION_RELATIVE -5.092816e-01
## DAYS_BIRTH 1.311621e-06
## DAYS_EMPLOYED 3.345873e-06
## OWN_CAR_AGE -4.561931e-04
## OCCUPATION_TYPEAccountants -4.950778e-02
## OCCUPATION_TYPECleaning staff -6.480265e-02
## OCCUPATION_TYPECooking staff -2.701925e-02
## OCCUPATION_TYPECore staff -3.135803e-02
## OCCUPATION_TYPEDrivers -1.804419e-02
## OCCUPATION_TYPEHigh skill tech staff -4.831588e-02
## OCCUPATION_TYPEHR staff -1.518850e-01
## OCCUPATION_TYPEIT staff -1.464076e-01
## OCCUPATION_TYPELaborers -2.520263e-02
## OCCUPATION_TYPELow-skill Laborers -4.202923e-02
## OCCUPATION_TYPEManagers -3.199947e-02
## OCCUPATION_TYPEMedicine staff -6.977675e-02
## OCCUPATION_TYPEPrivate service staff -6.610302e-02
## OCCUPATION_TYPERealty agents -3.941700e-02
## OCCUPATION_TYPESales staff -2.813218e-02
## OCCUPATION_TYPESecretaries -5.264054e-02
## OCCUPATION_TYPESecurity staff -3.010362e-03
## OCCUPATION_TYPEWaiters/barmen staff 7.778561e-03
## CNT_FAM_MEMBERS2 -2.756765e-02
## CNT_FAM_MEMBERS3 -3.446242e-02
## CNT_FAM_MEMBERS4 -4.123259e-02
## CNT_FAM_MEMBERS5 -2.748915e-02
## CNT_FAM_MEMBERS6 1.447404e-02
## CNT_FAM_MEMBERS7 -1.534942e-01
## ORGANIZATION_TYPEAgriculture -1.349900e-01
## ORGANIZATION_TYPEBank -1.354154e-01
## ORGANIZATION_TYPEBusiness Entity Type 1 -1.332057e-01
## ORGANIZATION_TYPEBusiness Entity Type 2 -9.284995e-02
## ORGANIZATION_TYPEBusiness Entity Type 3 -1.178262e-01
## ORGANIZATION_TYPECleaning -1.816591e-01
## ORGANIZATION_TYPEConstruction -1.101466e-01
## ORGANIZATION_TYPECulture -2.546274e-01
## ORGANIZATION_TYPEElectricity -1.735521e-01
## ORGANIZATION_TYPEEmergency -1.931145e-01
## ORGANIZATION_TYPEGovernment -1.283026e-01
## ORGANIZATION_TYPEHotel -2.253144e-01
## ORGANIZATION_TYPEHousing -1.314230e-01
## ORGANIZATION_TYPEIndustry: type 1 -1.912816e-01
## ORGANIZATION_TYPEIndustry: type 11 -1.235535e-01
## ORGANIZATION_TYPEIndustry: type 12 -1.005350e-01
## ORGANIZATION_TYPEIndustry: type 13 -5.198332e-01
## ORGANIZATION_TYPEIndustry: type 2 -2.695556e-01
## ORGANIZATION_TYPEIndustry: type 3 -9.385171e-02
## ORGANIZATION_TYPEIndustry: type 4 -1.401997e-01
## ORGANIZATION_TYPEIndustry: type 5 -1.610004e-01
## ORGANIZATION_TYPEIndustry: type 6 -3.485733e-01
## ORGANIZATION_TYPEIndustry: type 7 -1.251853e-01
## ORGANIZATION_TYPEIndustry: type 9 -1.049312e-01
## ORGANIZATION_TYPEInsurance -2.521474e-01
## ORGANIZATION_TYPEKindergarten -1.351527e-01
## ORGANIZATION_TYPELegal Services -1.715343e-01
## ORGANIZATION_TYPEMedicine -9.452424e-02
## ORGANIZATION_TYPEMilitary -1.759598e-01
## ORGANIZATION_TYPEMobile -2.966944e-01
## ORGANIZATION_TYPEOther -1.272777e-01
## ORGANIZATION_TYPEPolice -1.795750e-01
## ORGANIZATION_TYPEPostal -1.478992e-01
## ORGANIZATION_TYPERealtor 1.197700e-02
## ORGANIZATION_TYPEReligion -4.315237e-01
## ORGANIZATION_TYPERestaurant -1.621937e-01
## ORGANIZATION_TYPESchool -1.296596e-01
## ORGANIZATION_TYPESecurity -1.713341e-01
## ORGANIZATION_TYPESecurity Ministries -1.461711e-01
## ORGANIZATION_TYPESelf-employed -1.074924e-01
## ORGANIZATION_TYPEServices -1.616420e-01
## ORGANIZATION_TYPETelecom -1.970062e-01
## ORGANIZATION_TYPETrade: type 1 -1.867797e-01
## ORGANIZATION_TYPETrade: type 2 -1.465355e-01
## ORGANIZATION_TYPETrade: type 3 -1.506544e-01
## ORGANIZATION_TYPETrade: type 4 -5.003380e-01
## ORGANIZATION_TYPETrade: type 6 -1.779148e-01
## ORGANIZATION_TYPETrade: type 7 -9.356927e-02
## ORGANIZATION_TYPETransport: type 1 -3.712120e-01
## ORGANIZATION_TYPETransport: type 2 -6.270483e-02
## ORGANIZATION_TYPETransport: type 3 -1.187089e-02
## ORGANIZATION_TYPETransport: type 4 -1.686197e-01
## ORGANIZATION_TYPEUniversity -1.557948e-01
## ORGANIZATION_TYPEXNA -3.434401e+00
## AMR_REQ_CREDIT_BUREAU_SUM1 -1.371431e-02
## AMR_REQ_CREDIT_BUREAU_SUM2 -8.409579e-03
## AMR_REQ_CREDIT_BUREAU_SUM3 -7.853147e-03
## AMR_REQ_CREDIT_BUREAU_SUM4 -1.972689e-02
## AMR_REQ_CREDIT_BUREAU_SUM5 -3.374414e-02
## AMR_REQ_CREDIT_BUREAU_SUM6 -9.022792e-03
## AMR_REQ_CREDIT_BUREAU_SUM7 -2.141140e-02
## AMR_REQ_CREDIT_BUREAU_SUM8 4.524399e-03
## AMR_REQ_CREDIT_BUREAU_SUM9 -5.411993e-02
## AMR_REQ_CREDIT_BUREAU_SUM10 -1.060823e-01
## AMR_REQ_CREDIT_BUREAU_SUM11 -1.343153e-01
## AMR_REQ_CREDIT_BUREAU_SUM12 -2.486506e-01
## AMR_REQ_CREDIT_BUREAU_SUM13 -3.002701e-01
## AMR_REQ_CREDIT_BUREAU_SUM14 -3.701406e-01
## AMR_REQ_CREDIT_BUREAU_SUM15 -3.421588e-01
## AMR_REQ_CREDIT_BUREAU_SUM16 -5.653461e-01
## AMR_REQ_CREDIT_BUREAU_SUM17 -6.339209e-01
## AMR_REQ_CREDIT_BUREAU_SUM18 -4.341098e-01
## AMR_REQ_CREDIT_BUREAU_SUM19 -5.551390e-01
## AMR_REQ_CREDIT_BUREAU_SUM20 -5.528268e-01
## AMR_REQ_CREDIT_BUREAU_SUM28 -6.212837e-01
## 97.5 %
## (Intercept) 8.412018e-01
## CODE_GENDERM 5.793506e-02
## NAME_CONTRACT_TYPERevolving loans -3.133389e-02
## AMT_INCOME_TOTAL 2.145779e-08
## FLAG_OWN_CARY -1.839188e-02
## FLAG_OWN_REALTYY 1.737697e-02
## AMT_CREDIT 8.510751e-09
## AMT_ANNUITY 4.606605e-07
## NAME_TYPE_SUITEChildren 7.197754e-02
## NAME_TYPE_SUITEFamily 7.784861e-02
## NAME_TYPE_SUITEGroup of people 1.404451e-01
## NAME_TYPE_SUITEOther_A 8.186785e-02
## NAME_TYPE_SUITEOther_B 5.371884e-02
## NAME_TYPE_SUITESpouse, partner 8.129319e-02
## NAME_TYPE_SUITEUnaccompanied 7.134837e-02
## NAME_INCOME_TYPECommercial associate 4.548906e-01
## NAME_INCOME_TYPEPensioner 6.155963e-01
## NAME_INCOME_TYPEState servant 4.519130e-01
## NAME_INCOME_TYPEStudent 6.675666e-01
## NAME_INCOME_TYPEUnemployed 7.309430e-01
## NAME_INCOME_TYPEWorking 4.748840e-01
## NAME_EDUCATION_TYPEHigher education 2.767328e-01
## NAME_EDUCATION_TYPEIncomplete higher 2.958824e-01
## NAME_EDUCATION_TYPELower secondary 3.223308e-01
## NAME_EDUCATION_TYPESecondary / secondary special 2.951317e-01
## NAME_FAMILY_STATUSMarried 1.224180e-02
## NAME_FAMILY_STATUSSeparated 3.168740e-02
## NAME_FAMILY_STATUSSingle / not married 4.094842e-02
## NAME_FAMILY_STATUSWidow 5.715336e-02
## NAME_HOUSING_TYPEHouse / apartment 8.629808e-02
## NAME_HOUSING_TYPEMunicipal apartment 8.310490e-02
## NAME_HOUSING_TYPEOffice apartment 5.548787e-02
## NAME_HOUSING_TYPERented apartment 1.244752e-01
## NAME_HOUSING_TYPEWith parents 1.091197e-01
## REGION_POPULATION_RELATIVE 3.131685e-01
## DAYS_BIRTH 5.120578e-06
## DAYS_EMPLOYED 9.136180e-06
## OWN_CAR_AGE 6.636396e-04
## OCCUPATION_TYPEAccountants 2.023234e-02
## OCCUPATION_TYPECleaning staff 3.090617e-02
## OCCUPATION_TYPECooking staff 5.538443e-02
## OCCUPATION_TYPECore staff 2.055321e-02
## OCCUPATION_TYPEDrivers 3.860446e-02
## OCCUPATION_TYPEHigh skill tech staff 1.283258e-02
## OCCUPATION_TYPEHR staff 1.366754e-01
## OCCUPATION_TYPEIT staff 9.027604e-02
## OCCUPATION_TYPELaborers 1.556963e-02
## OCCUPATION_TYPELow-skill Laborers 9.566536e-02
## OCCUPATION_TYPEManagers 2.064583e-02
## OCCUPATION_TYPEMedicine staff 1.994669e-02
## OCCUPATION_TYPEPrivate service staff 5.334315e-02
## OCCUPATION_TYPERealty agents 1.832366e-01
## OCCUPATION_TYPESales staff 2.000032e-02
## OCCUPATION_TYPESecretaries 1.340962e-01
## OCCUPATION_TYPESecurity staff 8.470802e-02
## OCCUPATION_TYPEWaiters/barmen staff 1.926817e-01
## CNT_FAM_MEMBERS2 2.983188e-02
## CNT_FAM_MEMBERS3 2.982147e-02
## CNT_FAM_MEMBERS4 3.003458e-02
## CNT_FAM_MEMBERS5 9.552393e-02
## CNT_FAM_MEMBERS6 3.444218e-01
## CNT_FAM_MEMBERS7 3.316215e-01
## ORGANIZATION_TYPEAgriculture 1.437548e-01
## ORGANIZATION_TYPEBank 1.525561e-01
## ORGANIZATION_TYPEBusiness Entity Type 1 1.307764e-01
## ORGANIZATION_TYPEBusiness Entity Type 2 1.662319e-01
## ORGANIZATION_TYPEBusiness Entity Type 3 1.357323e-01
## ORGANIZATION_TYPECleaning 2.550805e-01
## ORGANIZATION_TYPEConstruction 1.527996e-01
## ORGANIZATION_TYPECulture 1.374006e-01
## ORGANIZATION_TYPEElectricity 1.422696e-01
## ORGANIZATION_TYPEEmergency 1.515682e-01
## ORGANIZATION_TYPEGovernment 1.306271e-01
## ORGANIZATION_TYPEHotel 9.841759e-02
## ORGANIZATION_TYPEHousing 1.423328e-01
## ORGANIZATION_TYPEIndustry: type 1 1.161458e-01
## ORGANIZATION_TYPEIndustry: type 11 1.547262e-01
## ORGANIZATION_TYPEIndustry: type 12 2.996820e-01
## ORGANIZATION_TYPEIndustry: type 13 2.786922e-01
## ORGANIZATION_TYPEIndustry: type 2 1.119602e-01
## ORGANIZATION_TYPEIndustry: type 3 1.772190e-01
## ORGANIZATION_TYPEIndustry: type 4 1.662886e-01
## ORGANIZATION_TYPEIndustry: type 5 1.870633e-01
## ORGANIZATION_TYPEIndustry: type 6 1.925000e-01
## ORGANIZATION_TYPEIndustry: type 7 1.760064e-01
## ORGANIZATION_TYPEIndustry: type 9 1.673743e-01
## ORGANIZATION_TYPEInsurance 1.053712e-01
## ORGANIZATION_TYPEKindergarten 1.278935e-01
## ORGANIZATION_TYPELegal Services 2.111750e-01
## ORGANIZATION_TYPEMedicine 1.687200e-01
## ORGANIZATION_TYPEMilitary 1.028410e-01
## ORGANIZATION_TYPEMobile 1.135685e-01
## ORGANIZATION_TYPEOther 1.295482e-01
## ORGANIZATION_TYPEPolice 1.041582e-01
## ORGANIZATION_TYPEPostal 1.353386e-01
## ORGANIZATION_TYPERealtor 4.219468e-01
## ORGANIZATION_TYPEReligion 3.656013e-01
## ORGANIZATION_TYPERestaurant 1.268437e-01
## ORGANIZATION_TYPESchool 1.314263e-01
## ORGANIZATION_TYPESecurity 1.095940e-01
## ORGANIZATION_TYPESecurity Ministries 1.401697e-01
## ORGANIZATION_TYPESelf-employed 1.468869e-01
## ORGANIZATION_TYPEServices 1.398134e-01
## ORGANIZATION_TYPETelecom 1.556940e-01
## ORGANIZATION_TYPETrade: type 1 2.229915e-01
## ORGANIZATION_TYPETrade: type 2 1.443654e-01
## ORGANIZATION_TYPETrade: type 3 1.200599e-01
## ORGANIZATION_TYPETrade: type 4 2.979220e-01
## ORGANIZATION_TYPETrade: type 6 1.596685e-01
## ORGANIZATION_TYPETrade: type 7 1.681001e-01
## ORGANIZATION_TYPETransport: type 1 2.211115e-01
## ORGANIZATION_TYPETransport: type 2 2.190883e-01
## ORGANIZATION_TYPETransport: type 3 2.887638e-01
## ORGANIZATION_TYPETransport: type 4 9.942510e-02
## ORGANIZATION_TYPEUniversity 1.550397e-01
## ORGANIZATION_TYPEXNA -1.047061e+00
## AMR_REQ_CREDIT_BUREAU_SUM1 2.209506e-02
## AMR_REQ_CREDIT_BUREAU_SUM2 3.025904e-02
## AMR_REQ_CREDIT_BUREAU_SUM3 3.363044e-02
## AMR_REQ_CREDIT_BUREAU_SUM4 2.596637e-02
## AMR_REQ_CREDIT_BUREAU_SUM5 1.971981e-02
## AMR_REQ_CREDIT_BUREAU_SUM6 5.909404e-02
## AMR_REQ_CREDIT_BUREAU_SUM7 5.317304e-02
## AMR_REQ_CREDIT_BUREAU_SUM8 1.123533e-01
## AMR_REQ_CREDIT_BUREAU_SUM9 8.795301e-02
## AMR_REQ_CREDIT_BUREAU_SUM10 6.909664e-02
## AMR_REQ_CREDIT_BUREAU_SUM11 1.404227e-01
## AMR_REQ_CREDIT_BUREAU_SUM12 1.658377e-01
## AMR_REQ_CREDIT_BUREAU_SUM13 1.533711e-01
## AMR_REQ_CREDIT_BUREAU_SUM14 3.352075e-01
## AMR_REQ_CREDIT_BUREAU_SUM15 1.760994e-01
## AMR_REQ_CREDIT_BUREAU_SUM16 5.099801e-01
## AMR_REQ_CREDIT_BUREAU_SUM17 4.351645e-01
## AMR_REQ_CREDIT_BUREAU_SUM18 2.872795e-01
## AMR_REQ_CREDIT_BUREAU_SUM19 5.215201e-01
## AMR_REQ_CREDIT_BUREAU_SUM20 4.516868e-01
## AMR_REQ_CREDIT_BUREAU_SUM28 3.977509e-01Use completed datasets to see AUCs for the fitted models
#Dataset 1
par(mfrow=c(1,1))
cd1 <- mice::complete(application_MI, 1)
reg_cd1 <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY
+ NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE
+ REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS
+ ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM,
family=binomial)
roc(cd1$TARGET, fitted(reg_cd1), plot=T, legacy.axes=T)
##
## Call:
## roc.default(response = cd1$TARGET, predictor = fitted(reg_cd1), plot = T, legacy.axes = T)
##
## Data: fitted(reg_cd1) in 9171 controls (cd1$TARGET 0) < 829 cases (cd1$TARGET 1).
## Area under the curve: 0.6981#Dataset 2
cd2 <- mice::complete(application_MI, 2)
reg_cd2 <- glm(data=cd2, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY
+ NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE
+ REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS
+ ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM,
family=binomial)
roc(cd2$TARGET, fitted(reg_cd2), plot=T, legacy.axes=T)
##
## Call:
## roc.default(response = cd2$TARGET, predictor = fitted(reg_cd2), plot = T, legacy.axes = T)
##
## Data: fitted(reg_cd2) in 9171 controls (cd2$TARGET 0) < 829 cases (cd2$TARGET 1).
## Area under the curve: 0.6993#Dataset 3
cd3 <- mice::complete(application_MI, 3)
reg_cd3 <- glm(data=cd3, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY
+ NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE
+ REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS
+ ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM,
family=binomial)
roc(cd3$TARGET, fitted(reg_cd3), plot=T, legacy.axes=T)
##
## Call:
## roc.default(response = cd3$TARGET, predictor = fitted(reg_cd3), plot = T, legacy.axes = T)
##
## Data: fitted(reg_cd3) in 9171 controls (cd3$TARGET 0) < 829 cases (cd3$TARGET 1).
## Area under the curve: 0.6986Model Diagnostics
#Dataset 1
cd1 <- mice::complete(application_MI, 1)
#Binned residual plots
rawresid1 = cd1$TARGET - fitted(reg_cd1)
#continuous variables
binnedplot(x=cd1$AMT_INCOME_TOTAL, y = rawresid1, xlab = "AMT_INCOME_TOTAL", ylab = "Residuals",
main = "Binned residuals versus AMT_INCOME_TOTAL")
binnedplot(x=cd1$AMT_CREDIT, y = rawresid1, xlab = "AMT_CREDIT", ylab = "Residuals",
main = "Binned residuals versus AMT_CREDIT")
binnedplot(x=cd1$AMT_ANNUITY, y = rawresid1, xlab = "AMT_ANNUITY", ylab = "Residuals",
main = "Binned residuals versus AMT_ANNUITY")
binnedplot(x=cd1$REGION_POPULATION_RELATIVE , y = rawresid1, xlab = "REGION_POPULATION_RELATIVE", ylab = "Residuals",
main = "Binned residuals versus REGION_POPULATION_RELATIVE")
binnedplot(x=cd1$DAYS_BIRTH , y = rawresid1, xlab = "DAYS_BIRTH", ylab = "Residuals",
main = "Binned residuals versus DAYS_BIRTH")
binnedplot(x=cd1$DAYS_EMPLOYED , y = rawresid1, xlab = "DAYS_EMPLOYED", ylab = "Residuals",
main = "Binned residuals versus DAYS_EMPLOYED")
binnedplot(x=cd1$OWN_CAR_AGE , y = rawresid1, xlab = "OWN_CAR_AGE", ylab = "Residuals",
main = "Binned residuals versus OWN_CAR_AGE")
binnedplot(x=as.numeric(cd1$AMR_REQ_CREDIT_BUREAU_SUM) , y = rawresid1, xlab = "OAMR_REQ_CREDIT_BUREAU_SUM", ylab = "Residuals", main = "Binned residuals versus AMR_REQ_CREDIT_BUREAU_SUM")
#DAYS_EMPLOYED
temp <- cd1[cd1$DAYS_EMPLOYED != 365243,]
binnedplot(x=temp$DAYS_EMPLOYED , y = rawresid1, xlab = "DAYS_EMPLOYED", ylab = "Residuals",
main = "Binned residuals versus DAYS_EMPLOYED")
Interaction Effect
There are several pairs of variables we would like to check for possible interaction effects.
NAME_INCOME_TYPEversusAMT_CREDITThe amount of credit a client is given is typically based on his income type. For example, state servants are considered more stable and banks typically give them higher amounts of credits. Therefore, we believe there could be an interaction effect between these two variables. However, after taking an F-test, we could conclude that the interaction effect is not significant and we should not incorporate into our model.
bwplot(as.factor(TARGET)~AMT_CREDIT|as.factor(NAME_INCOME_TYPE), data = cd1, ylab = "TARGET")
reg_cd1 <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM, family=binomial)
reg_int <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM + AMT_CREDIT*NAME_INCOME_TYPE, family=binomial)
anova(reg_cd1, reg_int, test = "Chisq")## Analysis of Deviance Table
##
## Model 1: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM
## Model 2: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM +
## AMT_CREDIT * NAME_INCOME_TYPE
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 9863 5318.2
## 2 9860 5316.5 3 1.6452 0.6492NAME_EDUCATION_TYPEversusAMT_INCOME_TOTALThe amount of income is typically associated with the highest education he or her received. Therefore, we try to see if there’s any interaction effect between these two variables. The plots below suggest that their might be an interaction effect, though not very clear. But the F-test indicates that we should not add an interaction term between these two variables in our regression model.
bwplot(as.factor(TARGET)~AMT_INCOME_TOTAL|as.factor(NAME_EDUCATION_TYPE), data = cd1, ylab = "TARGET")
reg_cd1 <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM, family=binomial)
reg_int <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM + AMT_INCOME_TOTAL*NAME_EDUCATION_TYPE, family=binomial)
anova(reg_cd1, reg_int, test = "Chisq")## Analysis of Deviance Table
##
## Model 1: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM
## Model 2: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM +
## AMT_INCOME_TOTAL * NAME_EDUCATION_TYPE
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 9863 5318.2
## 2 9859 5314.8 4 3.4027 0.4928OCCUPATION_TYPEversusAMT_INCOME_TOTALDifferent kinds of occupation generally have different levels of incomes. For example, high skill tech staffs typically have higher incomes than cleaning staffs. Hence, we think there might be an interaction effect. However, the F-test suggests that the interaction effect is not significant enough and we should not add it into our model.
bwplot(as.factor(TARGET)~AMT_INCOME_TOTAL|as.factor(OCCUPATION_TYPE), data = cd1, ylab = "TARGET")
reg_cd1 <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM, family=binomial)
reg_int <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM + AMT_INCOME_TOTAL*OCCUPATION_TYPE, family=binomial)
anova(reg_cd1, reg_int, test = "Chisq")## Analysis of Deviance Table
##
## Model 1: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM
## Model 2: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM +
## AMT_INCOME_TOTAL * OCCUPATION_TYPE
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 9863 5318.2
## 2 9845 5294.4 18 23.756 0.1632ORGANIZATION_TYPEversusAMT_CREDITFor most banks’ practices, banks generally offer different credit amounts for people work in different organization. For example, most banks offer Fortune 500 companies’ employees more credits. As a result, we think there might be an interaction effect between these two variables. The plots below also suggests that there could be a potential interaction term to be added. However, the F-test indicates that the interaction term is not significant enough to be added into our model.
bwplot(as.factor(TARGET)~AMT_CREDIT|as.factor(ORGANIZATION_TYPE), data = cd1, ylab = "TARGET")
reg_cd1 <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM, family=binomial)
reg_int <- glm(data=cd1, TARGET~CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL + FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY + NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE + NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE + DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE + CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM + AMT_CREDIT*ORGANIZATION_TYPE, family=binomial)
anova(reg_cd1, reg_int, test = "Chisq")## Analysis of Deviance Table
##
## Model 1: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM
## Model 2: TARGET ~ CODE_GENDER + NAME_CONTRACT_TYPE + AMT_INCOME_TOTAL +
## FLAG_OWN_CAR + FLAG_OWN_REALTY + AMT_CREDIT + AMT_ANNUITY +
## NAME_TYPE_SUITE + NAME_INCOME_TYPE + NAME_EDUCATION_TYPE +
## NAME_FAMILY_STATUS + NAME_HOUSING_TYPE + REGION_POPULATION_RELATIVE +
## DAYS_BIRTH + DAYS_EMPLOYED + OWN_CAR_AGE + OCCUPATION_TYPE +
## CNT_FAM_MEMBERS + ORGANIZATION_TYPE + AMR_REQ_CREDIT_BUREAU_SUM +
## AMT_CREDIT * ORGANIZATION_TYPE
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 9863 5318.2
## 2 9809 5276.7 54 41.466 0.8942